Home

quantiNEMO

1. for this special parameter an empty argument is not identical to a missing one overwrite integer default 0 This parameter specifies how present output is treated if there is a danger of overwriting 34 CHAPTER 4 SIMULATION 39 0 no If the output of the running simulation will overwrite present files quantiNEMO will ask the user how to proceed Do you want to overwrite all the files that use it y es n o s kip and will suspend the simulation until the user responds to the question 1 yes Present output is overwritten without asking filename string default simulation This name will be used as the base filename for all outputs The output file extensions are added to this base filename If a file is written on a replicate periodic basis the replicate number will be added between the base name and the extension so that the same file is not overwritten periodically The same is true concerning generation periodic files see section 2 4 The base name may include the special expansion character used to build filenames when sequential parameters are present in the input parameter file See the discussion on sequential parameters in section 2 7 2 logfile string default quantinemo log This is the file name including the extension for the log file in which the simulation logs are recorded This log file is stored next to the executable and records the main information of each
2. 1 0 factor 0 6 0 4 T T T T T T 0 0 0 2 0 4 0 6 0 8 1 0 population density N K In this figure the growth rate of the curve has a very high value of 1e default value of parameter dispersal_growth_rate This results in a sharp change of the factor from the lower asymptote parameter dispersal_k_min to the upper asymptote parameter dispersal k max Below the threshold parameter dispersal max growth i e when the population density is low no migration occurs since the factor is zero If the population size is larger than the half of the carrying capacity the patch sends emigrants at the set dispersal rate parameter dispersal_rate since the factor is 1 dispersal_k_min decimal temporal default 0 This parameter specifies the lower asymptote of the slope dispersal k max decimal temporal default 1 This parameter specifies the upper asymptote of the slope dispersal k max growth decimal temporal default 1 This parameter specifies the population density with the maximal change dispersal k growth rate decimal temporal default 1e This parameter specifies the slope of the curve The high default value of 1e implies that the slope is more or less vertically i e that the change between lower and higher asymptote is instantaneously Thus the parameter disper sal k max growth serves as threshold of the population density below which CHAPTER 3 LIFE CYCLE 33 the factor is set to the val
3. 1 Reflective boundaries The borders are reflective Dispersers from the border patches cannot move beyond the border Border cells have thus less cells connected to them and their dispersal probabilities to the adja cent patches are higher e g m for the 1D Stepping Stone model m 3 corners m 2 for the 2D Stepping Stone model with four adjacent cells and m 5 corners m 3 for the 2D Stepping Stone model with eight ad jacent cells No dispersers are lost 2 Absorbing boundaries Dispersers of the border patches are lost if they choose to move beyond the border The dispersal probabilities of a border patch are not modified dispersal lattice dims matrix This parameter allows to specify the length and width of the 2D Stepping Stone lattice only used when dispersal model is set to 3 The argument is an integer matrix with two values The first value stands for the number of rows and the second value for the number of columns The product of the two values results in the number of patches and thus must match the parameter patch number If the parameter is not set guantiNEMO assumes that the 2D Stepping Stone lattice is guadratic If this is not possible due to the number of patches an error is returned dispersal propagule prob decimal temporal default 1 This parameter is only used for the Propagule pool Island model parameter dispersal model set to 1 It specifies the probability that a migrant will move to the propagule
4. observed heterozygosity within patch i following Nei and Chesser 1983 computed for each patch expected heterozygosity within patch i following Nei and Chesser 1983 computed for each patch F statistics following Nei and Chesser 1983 q adlt off fstat q adlt off fst q adlt off fis q adlt off fit q adlt off fst pair global F gm global F s global Fr pairwise Fer between patch i and j all pairwise combina tions computed F statistics following Weir and Cockerham 1984 la adlt off fstat wc q adlt off fst we q adlt off fis we a adlt off fit wc q adlt off fst wc pair global F gm global F Is global Fr pairwise Fer between patch i and j all pairwise combina tions computed Table 6 1 continued on next page CHAPTER 6 QUANTITATIVE TRAITS Stat name Description Table 6 1 Summary statistics available for quantitative traits continued 75 Chapter 7 Neutral markers quantiNEMO also allows the simulation of neutral markers such as microsatellites or SNPs with different mutation models K allele and Stepwise Different types of neutral markers can be combined within the same simulation The initial allele frequencies can be defined for each population separately 7 1 Architecture ntrl loci integer This parameter specifies the number of neutral marker loci per individual This parameter is mandatory for the simulation of a neutral marker
5. Extinction probability of a patch at each generation Chapter 4 Simulation This section describes general parameters of a simulation generations integer Number of generations to perform per replicate This parameter is mandatory and has no default replicates integer default 1 Number of replicates to perform per simulation Replicates are identical sim ulations in terms of parametrization but their outcome may differ due to stochastic events folder string default simulation yyyy mm dd hh mm ss This parameter specifies the folder for the output The default argument of this parameter differs from the rules as the default folder name is dynamic i e comprises the start time and date of the simulation The default folder name is simulation with the date and time as suffix in the format yyyy mm dd hh mm ss Year Month Day Hour Minute Second This dynamic default folder name allows to store each simulation separately avoiding that previous outputs are overwritten If the parameter is set the passed argument will be used as folder name i e without any addition of the time In this case it may happen that the new output wants to overwrite previous outputs The parameter overwrite see below allows to specify the rules for overwriting other outputs If the output should be stored directly in the working directory this parameter has to be listed in the settings file followed by an empty argument
6. Migrant pool Island model If the dispersal rate is m and the number of patches is np the probability to disperse to any n 1 non natal patch is rel while the probability to stay at home is 1 m 1 Propagule pool Island model In that modified version of the Island model a proportion of emigrants from a patch parameter dispersal_propagule_prob p disperse to the same non natal patch This propagule patch varies among patches and is reassigned at each generation Each offspring of a patch has a probability my to migrate to this propagule patch With probability ae it will disperse to any patch but its natal or propagule patch With a probability of 1 m it will stay at home 2 1D Stepping Stone model In the one dimensional Stepping Stone model patches are placed on a line and migrants can only move to one of the two adjacent patches If the dispersal rate is m the probability to disperse to one of the adjacent patches is m 2 while the probability to stay at home is 1 m The parameter dispersal_border_model allows to specify how to treat the border patches 3 2D Stepping Stone model In the two dimensional Stepping Stone model patches are placed on a grid or lattice and migrants can move to 4 or 8 adjacent patches set by the dispersal lattice range parameter below If the dispersal rate is m the probability to disperse to one of the ad jacent patches is m 4 or m 8 depending on the the parameter disper sal lattice ran
7. 2 1 7752 2 223 3 117 0 3409 2 003 2 1 17 2 3 2 9 2 1 000 2 0 2081 2 803 0 146 0 456 5 137 2 1 11 1 8 1 9 1 0 678 The first line contains the number of patches 2 patches here and the number of traits 5 The next five lines contain the five trait names The following lines contain the individual s info one individual per line The first number is the patch number of the individual followed by the genotypic value for each trait Each line ends with six columns consisting supplementary information on the individual see above if the parameter quanti_save_geno_value is set to 2 quanti_geno_value dir string default This parameter allows to specify the subdirectory where genotypic values are stored This directory has to be specified relative to the simulation folder pa rameter folder and may also contain subdirectories If not specified default the output is stored in the simulation folder parameter folder guanti geno value logtime integer temporal default 1 This parameter specifies the time interval of the genotypic value output Since the parameter may change over time the output may be generated at any generation quanti geno value script string default 7 It is possible to launch a script just after the genotypic value file is generated CHAPTER 6 QUANTITATIVE TRAITS 70 The argument of the parameter is the file name of the script The name of the genotypic value file is p
8. directional selection growth rate 2 8 3 OU DUO ND EE Why is this not the case Assume that all guantitative traits are under directional selection Thus the matrix of the growth rate has to be repeated leading to the following full matrix 414 5 6 2 3 7 2 8 3 4 5 6 2 3 7 This matrix expansion leads to a warning indicating that the number of rows is not an entire subset of the number of guantitative traits Based on this matrix it is now obvious that the growth rate of the third guantitative trait thus the first trait under directional selection has the growth rates 2 8 3 and not 4 5 6 5 3 Summary statistics The summary statistics listed in the table below are available for demography of the populations The column Stat name contains the name of the summary statistic used to specify which summary statistics are computed parameter stat for details see section 3 2 These names appear also in the output file The column Description contains a short description of the summary statistics Some of the summary statistics are available for adults and offspring indicated by adit off To obtain a certain summary statistic for adults the prefix adlt has to be added to the summary statistic name e g adlt allnb respectively the prefix off to obtain the summary statistic for offspring e g off allnb Some of the summary statistics are computed for each patch separately indicated by computed for each patch in the ta
9. 1 female the ID of the individual the ID of the mother the ID of the father and the fitness of the individual The ID is a unique identifier for each individual of a simulation in the format 345_23 meaning that this is the 345th individual born in patch 23 The IDs of the individual the mother and the father allow to extract pedigree informations if the CHAPTER 7 NEUTRAL MARKERS 84 output is stored for each generation and also to investigate the migration behavior of the individual and its parents An example of such a file ntrl_save_genotype is set to 2 5 4 20 2 loc 1 loc 2 loc 3 loc 4 1 1415 1019 2002 0820 1 1 10 1 1 1 0 1 0 345 1 0814 0219 2002 2020 1 1 11 1 8_1 2 4 0 334 1 0808 0217 1902 0820 1 1 12_1 5 38 5 1 0 123 5 1004 0917 1404 1007 1 1 16 5 9 5 3 2 0 999 5 2017 1010 2013 1812 1 0 17 5 3 2 9 2 1 000 5 2017 1008 2013 1811 1 1 11 4 82 9 2 0 678 The first line contains the number of patches 5 patches here the number of loci 5 the highest possible allele index 20 and the number of digits used to write each allele 2 The next four lines contain the locus names The following lines contain the individual s info one individual per line The first number is the patch number of the individual followed by the genotype Each column represents a locus and the first half of the locus represent the first allele index while the second half of the locus the second allele at the give
10. By default if none of the following parameters are set individuals do not disperse among patches dispersal_rate dispersal_rate_fem dispersal_rate_mal decimal matrix temporal default 0 These parameters allow to set the emigration rate If the argument is a single value the dispersal model used depends on the other parameters of this section But it is also possible to specify the dispersal rate explicitly between each pair of patches for both directions if the argument is a matrix A dispersal matrix has precedence over all other dispersal settings The matrix must be patch number x patch number in dimensions Each d element of this matrix is the dispersal probability from patch i to patch j where i specifies the row and 7 the column of the matrix Consequently the values in a row must sum up to 1 The dispersal rates can either be specified for both sexes in general or for each sex separately If the dispersal rates are sex specific the dispersal rates for both sexes have to be specified and they have to be in the same format CHAPTER 3 LIFE CYCLE 29 matrix or a single dispersal rate Sex specific dispersal rates have precedence over a general dispersal rate Note that the dispersal matrix has to be fully specified i e the matrix is not adjusted to the number of patches as for other parameters dispersal model 0 1 2 3 default 0 The following dispersal models can be specified if the dispersal rate is a single rate 0
11. Line by line the index of the columns to be read have to be declared Thereby a keyword for the specific setting is followed by the column index the ordering starts with 1 The following column keywords are available col locus This keyword specifies the column containing the locus index If this column is not declared in the file information box quantiNEMO will use the same settings for all loci In this case the length of the table must meet the number of alleles ntrl_all If this keyword is declared the length of the table must meet the number of alleles times the number of loci ntrl loci ntrl_all col allele This keyword specifies the column containing the allele index This column is mandatory The index of the allele goes from 1 to ntrl_loci col mut freq This keyword specifies the column containing the mutation probabilities i e the probability to mutate to this allele when a mu tation occurs The behavior of this mutation probability depends on the mutation model see parameter ntrl_mutation_model col_ini_freq This keyword specifies the column containing the initial frequen cies of the alleles This column allows to explicitly set the allele frequen cies at the start of a simulation The frequencies can be set for each patch separately using a matrix In the example above individuals of the first population are initially fixed for the allele 3 at the first locus as well as at the second locus In the second population all
12. alleles have the same initial frequency of 0 2 Note that the matrix is adjusted in length if the number of populations does not correspond to the length of the matrix CHAPTER 7 NEUTRAL MARKERS 78 If this column is not given the initial allele frequencies are set globally depending on the parameter ntrl_ini_allele model Note that as in all input files for quantiNEMO it is possible to define comments also in the file information box using the hash character or any text PE 7 2 Mutation lt Does the allele mutate gt gt No Yes Mutation model KAM SEM Me the mutation _ probabilities defined explicilty we be the mutation 7 _ probabilities defined gt Ss explicilty 27 a Yes 7 No a SS DR a Allele as drawn following the given probabilities Needed parameters nil mutation_model 0 ntrl_mutation_rate ntrl_allelic_file Allele a drawn in an uniform distribution Needed parameters ntrl_mutation_model 0 ntri mutation rate 4 with cof_mut_freg ya Step size is randomly drawn following the given probabilities Needed parameters ntrl_mutation_model 1 ntrlmutation_rate ntrl_allelic_file f A direction or is randomly drawn Needed parameters ntrl_mutation_model 1 ntrl mutation rate t with co mut freg 1 Brew T Box step Brew But Figure 7 1 Schematic representation of the mutation
13. before the extension An example of such a file name is simulation_g05r4 gen quanti_save_geno_value 0 1 2 default 0 This parameter specifies the output of the phenotype 0 None No output is generated 1 FSTAT The output contains the phenotypes in the FSTAT like format Goudet 1995 CHAPTER 6 QUANTITATIVE TRAITS 69 2 FSTAT extended Same as point 1 but the file contain the following six additional columns the age class 1 offspring 2 adult the sex 0 male 1 female the ID of the individual the ID of the mother the ID of the father and the fitness of the individual The ID is a unique identifier for each individual of a simulation in the format 345_23 meaning that this is the 345th individual born in patch 23 The IDs of the individual the mother and the father allow to extract pedigree informations if the output is stored for each generation and also to investigate the migration behavior of the individual and its parents An example of such a file quanti_save_geno_value is set to 2 2 5 genotypic_value_trait l genotypic_value_trait 2 genotypic_value_trait 3 genotypic_value_trait 4 genotypic_value_trait 5 1 0 0493 3 203 2 441 0 0683 3 199 2 1 10 1 1 1 0 1 0 345 1 0 4924 3 803 0 869 2 002 2 594 2 1 11_1 8 1 2 2 0 334 1 2 2342 2 931 0 725 0 750 0 698 2 1 12 1 5 2 5 1 0 123 2 0 8623 0 6525 0 857 1 7483 4 194 2 1 16 2 9 2 3 2 0 999
14. by quantiNEMO For example the first sequential parameter has to be called as name_ 1_4K and not as name_ 14K If the filename for the example above was set to sim_ 2pop_ lind the following base names would be generated alphabetically patch capacity comes before pach_number filename 1 simulation sim_l0pop 5ind 2 simulation sim_10pop 10ind 3 simulation sim 10pop 20ind 1 simulation sim_50pop 5ind 5 simulation sim_50pop 10ind 6 simulation sim_50pop 20ind Chapter 3 Life Cycle quantiNEMO is a discrete generation based simulator This means that a single individual undergoes only once the life cycle and that the generations are not over lapping Depending on the parameterization in the settings file some events may be skipped The life cycle has a fixed order of events A simulation starts with Breeding i e only adults are present at the initialization of the simulation The life cycle is repeated for each generation i e the life cycle event Breeding follows the life cycle event Extinction Breeding Adults mate and may produce offspring Selection acts at this stage Statistics Adults and juveniles are present It is the stage were summary statistics may be computed for adults and juveniles Outputs Adults and juveniles are still present Genotypes and or phenotypes may be dumped to files for adults and or juveniles Aging It is the event where t
15. components including genetic variance estimates quantitative trait analysis e g Qsr and F statistics for all types of loci neutral and QTL Most of the summary statistics are available for juveniles and adults The summary statistics can be computed for any generation during the simulation quantiNEMO can also produce files with the raw genetic data The genotypes at all loci can be dumped to file in the FSTAT format Goudet 1995 Phenotypes as well as the additive dominance and epistatic effect values can be written to a file and then analyzed with any population or quantitative genetic software e g to get patterns of differentiation study linkage disequilibrium or scan for QTLs We are currently developing an R package to carry out the most common analyses from a simulation Chapter 2 Using quantiN EMO This chapter explains how to use quantiNEMO starting from the basics 2 1 Installation Executables of quantiN EMO for several operating systems can be downloaded from the web site http ww unil ch popgen softwares quantinemo The executa bles are stand alones meaning that quantiNEMO does not require an installation After downloading the compressed file with the executable corresponding to your operating system simply extract it to a folder of your choice The compressed file includes the executable for your operating system the user manual and also an example settings file 2 2 Launching quantiNEMO A settings file
16. for Win dows Linux and Mac the source code a detailed user s manual and also a syn tax highlighting definition to edit the settings file with the shareware TextPad http www textpad com All downloads are freely available under the terms of the GNU General Public License 1 3 Technical quantiNEMO is a console program and is coded in standard C using an object oriented approach This allows compiling quantiNEMO on any computer platform which supports standard C compilation There is no limit on the number of pop ulations individuals genes etc that quantiNEMO can handle apart from the avail able hardware capacities CPU and memory quantiNEMO was optimized for high computation efficiency in particular for large simulations on clusters quantiNEMO 1 CHAPTER 1 INTRODUCTION 2 is built on the evolutionary and population genetics programming framework NEMO Guillaume and Rougemont 2006 with well developed demographic models The demographic models of NEMO were kept although several of these functionalities were re coded respectively adapted to the new functionalities of quantiNEMO 1 4 License quantiNEMO is free software you can redistribute it and or modify it under the terms of the GNU General Public License as published by the Free Software Foun dation either version 3 of the License or at your option any later version quan tiNEMO is distributed in the hope that it will be useful but WITHOUT ANY WAR RAN
17. genotypes genotypic and phenotypic values consider all patches The parameter sampled_patches al lows to make a selection among the patches In this case the summary statistics are only computed among the sampled patches and genotypes genotypic and phenotypic values are only dumped to file for these selected patches This parameter thus allows for example to investigate different sampling schemes There are different methods to define the sampled patches matrix Using a one dimensional matrix as argument allows to define explic itly the patches to sample Please note that for this parameter a matrix has to be defined fully i e all patches to sample have to be listed in the matrix and a single number is not expanded to a matrix number If the argument is a single number the patches to sample are drawn randomly In this case the passed number specifies the total number of patches to sample Patches are drawn randomly for each replicate but remain the same during a simulation CHAPTER 4 SIMULATION 37 0 This is the default value Also a single number but in this case all patches are sampled Chapter 5 Metapopulation This section describes general parameters concerning the metapopulation patch number integer default 1 This parameter specifies the number of patches in the metapopulation 5 1 Population sizes This section allows to specify the carrying capacities of the patches and the initial populations patch c
18. if they have the same specifications or for each locus separately The dominance file has a similar format as the allelic file Dominance file FILE_INFO col_locus 1 col_allelel 2 col_allele2 3 col_dominance 4 locus allelel allele2 value 1 1 2 0 298 1 1 3 0 435 1 1 4 0 224 1 2 3 0 104 1 2 4 0 974 1 3 4 0 808 The file has to start with a file information box FILE_INFO This box con tains informations about the structure of the following table and thus allowing flexibility in the format of the table For example the order of the columns in the table may vary or some columns may be ignored The file information box starts with the key word FILE INFO This key word is followed by brackets within which the user has to specify the contents of the columns to be considered by quantiNEMO Each column is specified by a pair consisting of a key word e g col_locus followed by the column number the ordering starts with 1 Each column definition has to be on a new line The following column keywords are available col locus This keyword specifies the column containing the locus index If this column is not declared in the file information box quantiNEMO will use the same settings dominance effect and or fitness factor for all loci col_allelel This keyword specifies the column containing the index of the allele with the smaller allelic effect This column is mandatory The index of the allele goes f
19. is 0 default value all loci have the same mutation rate defined by the parameter quanti_mutation rate The parameter quanti_mutation shape is ignored if the mutation rates parameter quanti_mutation rate are explicitly defined for each locus by a matrix 6 3 Initial genotypes There are several methods to set the allele frequencies or even the genotypes of the in dividuals at the start of a simulation initialisation The genotypes of the individu als may be set using an FSTAT file Goudet 1995 parameter guanti ini genotypes If such a FSTAT file is not present the genotypes are randomly drawn following the explicitly set allele frequencies in the allelic file see parameter quanti_allelic_file and especially column keyword col_ini_freq If the allele frequencies are not set explicitly in the allelic file the initialization is performed following the parameter guanti ini allele model quanti ini genotypes string default 7 This parameter allows to specify a name of an FSTAT file Goudet 1995 containing the initial genotypes of the individuals for each population If such a file is present the initialization of the metapopulation is done solely by this file ignoring the parameters quanti ini allele model and patch ini size A single FSTAT file is needed for all quantitative traits together containing the total number of loci of all quantitative traits together Note that quan tiNEMO allows to output an appropriate file for any ge
20. it is possible to pass an array of seeds up to 624 to quantiNEMO if specified as a one dimensional matrix If the seed is not set i e is set to 1 the random generator is initialized by the time i e a matrix with two seeds is used Thereby the first number is the time in seconds since 1 1 1970 function time NULL in C and the second number is the number of clock ticks elapsed since quantiNEMO started function clock in C seed 2354135 20234510 30345120 456300 50363450 postexec_script string default genetic_map_output 0 1 default 0 quantiNEMO allows simulating linked neutral markers and QTLs The sec tions 6 6 for QTLs and 7 5 for neutral loci describe how to specify the corresponding genetic map The parameter genetic_map_output allows to spec ify if the underlying genetic map is dumped to file or not 0 output The genetic map is dumped to file for every replicate if the genetic map was simulated i e if at least one of the parameters quanti_loci_positions quanti_loci_positions_random ntrl_loci_positions and ntrl_loci_positions_random are set The output is named genetic_map txt and is stored in the sim ulation folder If multiple replicates are simulated a replicate counter example _r3 is inserted before the extension 1 no output The genetic map is never dumped to file sampled_patches integer matrix default 0 By default summary statistics and all other outputs
21. ntrl all 1 to 256 default 255 This parameter specifies the maximal number of alleles per locus same number for each locus ntrl_allelic file string default This parameter allows to pass the name of a file containing allele informations such as the initial allele frequencies and or the mutation probability to an allele The number of alleles and loci has to be in line with the parameters ntrl loci and ntrl all The information can be set globally for all loci if they have the same specifications or for each locus separately The allelic file for neutral markers has the same format as the allelic file for quantitative traits but does not allow to specify the allelic effects Allelic file THR RAR FILE_INFO col_locus 1 col_allele 2 col mut freg 3 76 CHAPTER 7 NEUTRAL MARKERS TT col_ini_freq 4 locus allele mut freg ini_freq 1 1 0 20 10 0 2 1 2 0 25 10 0 2 1 3 0 20 1 0 2 1 4 0 20 0 0 2 1 5 0 20 0 0 2 2 1 0 20 0 0 2 2 2 0 25 0 0 2 2 3 0 20 1 0 2 2 4 0 20 0 0 2 2 5 0 20 10 0 2 The file has to start with a file information box FILE INFO This box contains the information of the structure of the following table allowing a flexible structure of the table For example the order of the columns in the table may vary or columns may be ignored The file information box starts with the key word FILE INFO and the information is enclosed by brackets P
22. of one dimensional matrices with each value of the matrix corresponding to one population The lines of the matrix are enclosed in braces In the example above indi viduals of the first population are initially fixed for the allele 3 at the first locus as well as at the second locus In the second population all alleles have the same initial frequency of 0 2 Note that the matrix is adjusted in length if the number of populations does not correspond to the length of the matrix If this column is not given the initial allele frequencies are set globally depending on the parameter quanti_ini_allele model CHAPTER 6 QUANTITATIVE TRAITS 52 Note that as in all input files for quantiNEMO it is possible to add comments also in the file information box using the hash character or any text quanti_allelic_var decimal matrix default 1 This parameter allows to specify the variance of the normal distribution from where the allelic effects are drawn randomly The normal distribution is cen tered around 0 mean phenotype is by definition 0 By using a matrix it is possible to define different variances of the normal distributions for each locus i e specifying different contributions of the QTLs to the trait This parameter is taken into account only if the allelic effects are not defined explicitly by the allelic file 6 1 2 Dominance effects It is possible to simulate dominance effects between alleles at a loc
23. offspring oz is logistically regulated Beverton and Holt 1957 and depends therefore on the carrying capacity K and on the pa rameter growth_rate r 5 stochastic logistic regulation No Poisson ver Same as point 4 but the computation of the total number of offspring has a stochastic component Models and their specific parameters model additional parameters 0 carrying capacity 1 keep number 2 fecundity mean fecundity 3 fecundity stochastic mean fecundity 4 logistic regulation growth rate 5 stochastic logistic regulation growth rate growth rate decimal This parameter is mandatory and only used if logistic regulation is used to CHAPTER 3 LIFE CYCLE 24 specify the number of offspring i e parameter breed model set to 4 or 5 It specifies the growth rate of the population mean fecundity integer This parameter specifies the mean female fecundity The parameter is manda tory and only used if the fecundity of the female specifies the number of offspring i e parameter breed_model set to 2 or 3 mating system 0 1 2 3 4 default 0 Five general mating systems are implemented in quantiNEMO The assign ment of the parents to the offspring is random depending on the fitness of the local parents if no selection acts all individuals have a fitness of 1 Thereby adults with a higher fitness have on average a higher reproductive success 0 random mating hermaphrodite For ea
24. quanti_mutation_var Depending on the mutation model parameter quanti_mutation_model the mutant effect is the effect of the drawn allele model 0 or the effect of the drawn allele is added to the current allelic effect to get the mutant effect model 1 Using the allelic file see section 6 1 1 it is possible to specify for each allele its effect and the probability to mutate to this allele given that there is a mutation A minimal definition for mutations requires the setting of a common mutation rate parameter quanti_mutation_rate has a single value In this case all loci have the same mutation rate and the mutation model is RMM quanti_mutation_model 0 1 default 0 This parameter allows to specify the mutation model The following mutation models are available 0 RMM Random Mutation Model At a given mutation a new allele is drawn randomly Anew Amut Where anew is the effect of the new allele and amut is the effect of the drawn allele The probability to mutate to a certain allele depends on its probability to mutate to it given that there is a mutation These probabilities and also the effects of the alleles can explicitly be set by the allelic file see section 6 1 1 If this is not the case quantiNEMO allocates the allelic effects and the probabilities to mutate to the alleles given that there is a mutation automatically The effects of the alleles are regularly spaced between 60 and 60 if there are more than 5 allele
25. the loci are assumed to be unlinked and independent ntrl loci positions matrix default gt For each trait this parameter allows to specify explicitly the loci position in centi Morgans from the beginning of the chromosome The brackets separate the chromosomes Within a chromosome loci are separated by a space Note that chromosomes may contain different numbers of loci but that the total number of loci must correspond to the parameter ntrl_loci It is possible to skip chromosomes if they don t contain loci see section 2 3 5 4 ntrl_loci 11 ntrl_loci_positions 1 10 3 10 40 60 80 100 10 40 60 80 100 In the example above the neutral marker is defined by 11 loci located on three of at least four chromosomes The first chromosome contains a single locus at position 10 cM No loci are located on the second chromosome The third and fourth chromosomes have 5 loci at positions 10 40 60 80 and 100 cM CHAPTER 7 NEUTRAL MARKERS 83 ntrl_loci_positions_random matrix default gt This parameter allows to specify the number of chromosomes and their lengths The loci are then randomly positioned on the chromosomes The matrix con tains the length of each chromosome in cM It is possible to jump chromo somes if they don t contain loci see section 2 3 5 4 ntrl loci 11 ntrl_locus_positions_random 1 100 13 100 100 11 loci of this type of neutral marker are locat
26. this latter case an additional statistic named alive rpl will be added which contains the number of alive replicates i e the number of simulations where the populations did not get extinct If no summary statistics are computed this event is skipped stat_save 0 1 2 3 4 5 default 0 This parameter specifies if the summary statistics should be computed and how they should be dumped to file The summary statistics may be dumped to file for each specified generation and replicate separately file generic name stats txt or summary statistics may be summed up across replicates by their mean file generic name stats txt and their variance file generic name var txt 0 All Output includes all types of summary statistic files generic name_stats txt generic_name_mean txt and generic_name_var txt 1 Detailed Output includes only the file containing the summary statistics for each replicate separately file generic name stats txt 2 Summed up Output includes the files containing the summary statistics summed up by their mean and variance across replicates files generic name mean txt and generic_name_var txt 3 Mean Output includes only the file containing the summary statistics summed up by their mean across replicates file generic_name_mean txt 4 Variance Output includes only the file containing the summary statistics summed up by their variance acros
27. variance of the normal distribution by which the phenotype with maximal growth varies at each generation e g annual fluctuations of the mean temperature By default the local phenotype with maximal growth does not vary patch dir sel symmetry var decimal matrix default 0 This parameter specifies the variance of the normal distribution by which the symmetry of the curve varies at each generation e g annual fluctuations of the mean temperature By default the symmetry of the slope does not vary Example patch dir sel growth rate 1 patch dir sel max growth 0 patch dir sel symmetry 1 In this example the selection pressure for all patches and guantitative traits are iden tical and set to the default values The specified directional selection pressure favors larger phenotypes parameter patch dir sel growth rate is positive This means that individuals with larger phenotypes have on average higher fitnesses and thus higher reproductive successes 5 2 3 Multiple traits with varying types of selection A special case arises if multiple quantitative traits are simulated with varying types of selection The problem is how to specify the individual selection pressures This is a more technical problem which I would like to illustrate here First quantiNEMO investigates which types of selection will be simulated based on the settings file Then quantiNEMO sets for each simulated type of selection the corresponding pa rame
28. 007 1 1 16 5 9 5 3_2 0 999 5 2017 1010 2013 1812 1 0 17 5 3 2 9 2 1 000 5 2017 1008 2013 1811 1 1 11 4 82 9 2 0 678 The first line contains the number of patches 5 patches here the number of loci 4 the highest possible allele index 20 and the number of digits used to write each allele 2 The next four lines contain the locus names The following lines contain the individual s info one individual per line The first number is the patch number of the individual followed by the genotype Each column represents a locus and the first half of the locus first 2 digits repre sents the first allele index while the second half of the locus last 2 digits the second allele at the given locus As in this example we are using two digits per allele the first two digits of a locus genotype number are the first allele e g allele 14 for the first allele of the first locus of the first individual while the two next digits are the second allele e g allele 15 for the second allele of the first locus of the first individual Each line ends with six columns consist ing supplementary information on the individual see above if the parameter quanti_save_genotype is set to 2 quanti genot dir string default This parameter allows to specify the subdirectory where the genotypes are CHAPTER 6 QUANTITATIVE TRAITS 68 stored This directory has to be specified relative to the simulation folder pa rameter folder and may also c
29. 4 simulation mysim 4 5 simulation mysim 5 6 simulation mysim 6 However the base file name can also be individualized by the user using expansion characters in the base file name The expansion characters allow to incorporate the changing argument value of the sequential parameters in the base file name For this the expansion character has to be followed by a number which corre sponds to the rank of the sequential parameter Note that the rank corresponds to the alphabetical order of the sequential parameter names starting with 1 for the first sequential parameter The expansion characters and the rank numbers are then automatically replaced by the corresponding arguemnt values used in the sim ulation Expansion characters can be applied to all types of arguments however arguments of matrices and temporal parameters are not replaced by the argument value but by the rank number of the argument how they appear in the settings file after the parameter name due to their big size So each sequential parameter can be addressed by its rank allowing to build separate filenames If not all sequen tial parameters are addressed by the filename i e if the base name is not unique for each simulation the rank of the simualtion is added as suffix to the filename to avoid overwritting the output Note that after the rank number which may consists of several digits a character must follow which cannot be interpreted as a number
30. 54 quanti environmental model 56 quanti environmental proportion 57 quanti_epistatic file 54 quanti_epistatic_var 56 quanti_geno_value_age 70 quanti_geno_value_dir 69 quanti_geno_value logtime 69 quanti_geno_value_script 69 quanti_geno_value_sex 70 quanti_genot_age 68 quanti_genot_dir 67 quanti genot logtime 68 quanti_genot_script 68 quanti_genot_sex 68 quanti_heritability 57 quanti_ini_allele_model 63 quanti ini genotypes 62 quanti_loci 50 quanti_loci_positions 65 quanti_loci_positions_random 66 quanti_mutation_model 59 quanti mutation rate 61 91 quanti_mutation_shape 61 quanti_nb_trait 63 quanti output 72 quanti_phenot_age 72 quanti_phenot_dir 71 quanti_phenot logtime 71 quanti_phenot_script 71 quanti phenot sex 71 quanti_save_geno_value 68 quanti_save_genotype 66 quanti_save_phenotype 70 quanti_selection model 64 dguanti va model 58 regulation model adults 33 regulation model offspring 28 replicates 34 sampled patches 36 seed 35 sex ratio 25 stat 27 stat dir 27 stat log time 26 stat_save 26
31. CHAPTER 6 QUANTITATIVE TRAITS 74 Stat name Description q adlt off thetaMM mean within patch within males coancestry q adlt off thetaFM mean within patch between sexes coancestry q adlt off coa fsib mean coancestry within full siblings q adlt off coa phsib mean coancestry within paternal half siblings q adlt off coa mhsib mean coancestry within maternal half siblings q adlt off coa nsib mean coancestry within non siblings q adlt off theta_p mean coancestry within patch 7 computed for each patch q adlt off alpha_pair mean coancestry between patch 7 and j all pairwise com binations computed Genetic diversity q adlt off gendiv q adlt off nbAll q adlt off meanAll q adlt off q adlt off q adlt off h q adlt hes d adlt off OA UU q adlt off nbAll_p q adlt off nbFixLoc p q adlt off ho_p q adlt off hs p nbFixLoc meanFixLoc mean number of alleles per locus across the entire metapop ulation mean number of alleles per locus and patch number of fixed loci across the entire metapopulation mean number of fixed loci per patches observed heterozygosity following Nei and Chesser 1983 expected heterozygosity following Nei and Chesser 1983 total expected heterozygosity following Nei and Chesser 1983 mean number of alleles per locus within patch 7 computed for each patch number of fixed loci within patch computed for each patch
32. EE EE NE OES EG N 5 22 Launching Qui o esse ee ee die OE RE RR OE EE N 5 Ad Input the settings ile is bia cb dba iia ada 6 RE MEES EES EE OE ara A 6 AD CON so sd EA a A AR 6 ER Aa RA T Ll DO ccoo rad aa a e E 7 ci A eco eane n egna na eT e a a a e a aG 8 Cte rr e 8 Dato Decimal a a E E A 8 E DINIE A 8 ka e or a EE a Be te aes ai 9 2 3 6 Temporal parameters 10 29 External TI MOE er EET DEER EER OE N 11 CONTENTS o A BL IVS EER N ee IA A 24 2 Naming convention o o ira a fee de 259 Blas RS Dle e socso ea ce BOER o AA A A 2060 UNO MMS e LL Hu Se BOER Dee EES tes OS GE RE AT Bek MON seca RA ee AR Ee A MEN 2 7 1 Multiple settings files 2 1 2 Sequential parameters 6 604 4048 ee wee a eG 3 Life Cycle ML Breme o de Aa A dd GP Mey A IEEE Det IED EER A Bok A Bch ah Hw e Meee it aos Se JA SUE EE EE ET EET HEET A be 3 5 Regulation offspring 2 2 one s esre us a e du RE due a 36 Diperal socre seda ee oe he sos DO dette 3 6 1 Density dependent dispersal rate ot Regulation cd aa raara pa DER RE GR A o DINO rr Dia du A A 4 Simulation 5 Metapopulation A o LL ER ee Bh EES ek RS EE EG 5 2 Selection pressure lt lt ee esos 5 2 1 Stabilizing selection ia oros eek ee ee ag 4 5 2 2 Directional selection 2 264654 sucres eee RE 5 2 3 Multiple traits with varying types of selection mee OU caia 6 Quant
33. INEMO 9 2 3 5 4 Matrix Matrices allow to pass several numbers integer or decimal to a parameter This may be necessary to specify carrying capacities see section 1D Matrix or to pass a dispersal matrix see section 2D Matrix to quantiNEMO Matrices are enclosed between curly brackets and numbers are separated by at least a space Matrices may be written on several lines and may also contain comments There is no a priori restriction on the size of the matrix 1D Matrix A one dimensional matrix vector consists of data in a single dimension There are three different ways to write a one dimensional matrix which are all equivalent patch_number 4 patch_capacity 20 10 20 10 patch capacity 120 10 20 10 patch_capacity 20 10 20 10 2D Matrix Some parameters need a second dimension for their argument A second dimension is obtained by enclosing the inner rows of the matrix again within curly brackets i patch number disp rate 4 0 0 0 0 S S AANO SS OD amp D OO S amp S D 00 EA 10 2 0 4 0 4 0 0 This example shows the pairwise dispersal matrix for 4 patches 4x4 Each row specifies a source patch from which emigrants emigrate Each column specifies the target patch receiving the immigrants The diagonal of the matrix specifies the proportion of individuals remaining in the natal patch Matrix length adjustment Usually mat
34. MO Each column is specified by a pair consisting of a key word e g col_locus followed by the column number the ordering starts with 1 Each column definition has to be on a new line The following column keywords are available col_locus This keyword specifies the column containing the locus index If this column is not declared in the file information box quantiNEMO will use the same settings for all loci In this case the number of lines of the table must be equal to quanti_all If the keyword col locus is declared the number of rows of the table must be equal to the number of alleles times the number of loci quanti_loci quanti_all col_allele This keyword specifies the column containing the allele index This column is mandatory The index of the allele goes from 1 to quanti_all col_allelic_value This keyword specifies the column containing the allelic ef fects If this column is not set the allelic effects will be drawn randomly from a normal distribution with the variance defined by the parameter quanti_all_effect_var col_mut_freq This keyword specifies the column containing the probability to mutate to this allele given that a mutation occurs col ini freq This keyword specifies the column containing the initial frequen cies of the alleles This column allows to explicitly set the allele frequen cies at the start of a simulation The frequencies can be set for each patch separately in which case the column consists in fact
35. TY without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE See the GNU General Public License for more details You should have received a copy of the GNU General Public License along with quantiNEMO If not see http ww gnu org licenses 1 5 Acknowledgments We are grateful to Yves Rousselle Patrick Meirmans Christine Grossen Claire Mouton Olivier Blaser Ilkka Kronholm YiJian Huang and Patrick Flight These persons helped us to improve quantiNEMO by reporting bugs and or great discus sions 1 6 Main features quantiNEMO consists in several simulation components which may be easily ex tended The simulation components with their corresponding parameters are de scribed in more detail in the rest of this manual Quantitative traits quantiNEMO allows the simulation of one to multiple quantitative traits each having its own specifications Each trait is defined by one to many loci each with up to 256 alleles The allelic effects at each locus can be drawn from a normal distribution or can be set explicitly Mutations are implemented with several models The trait determinism can be purely additive or include dominance and or epistatic interactions among loci Environmental effects can also be set in different ways Neutral markers quantiNEMO also allows the simulation of neutral markers such as microsatellites CHAPTER 1 INTRODUCTION 3 or SNPs with different mutation models K Allele Stepwis
36. age is output Phenotypes are only computed at the reproduction stage thus juveniles have no phenotypes re spectively are indicated as NaN not a number in the output This parameter is only present for uniformity reasons to obtain the same output format 0 Adults Output includes only adult phenotypes 1 Juveniles Output includes only juvenile phenotypes respectively NaNs 2 Both Output includes juveniles and adults phenotypes 6 7 4 Architecture Depending on the definition of the architecture of the quantitative trait it is possible to obtain the allelic file the dominance file and or the epistatic file All three files follow the structure of the homonymous input files except that for the input all possible combinations have to be present while in the output only the used com binations are output Therefore the output files of allelic dominance and epistatic cannot always be used as input files The files have the names allelic_values txt dominance values txt and epistatic_values txt and are stored in the simulation folder parameter folder If several replicates are simulated the files are generated for each replicate and a replicate counter e g _r4 is inserted before the extension quanti_output 0 1 default 0 0 None The files are not generated 1 Output The allelic the dominance and or the epistatic files are stored in the simulation folder parameter folder if they were used during the si
37. ain the same number of columns Other parameters such as the parameter quanti_loci_positions allow to skip a row i e some rows do not contain any data A row can be skipped by explicitly indicating the rank of the row just after the beginning of the row followed by a colon The ranking starts with 1 If a row has no explicit rank it is assumed to follow the preceding row Here is an example for the genetic map of a quantitative trait guanti loci 11 quanti_loci_positions 1 10 13 10 40 60 80 100 10 40 60 80 100 In this example the quantitative trait is defined by 11 loci located on three out of four chromosomes The first chromosome contains a single locus at position 10 cM No locus is located on the second chromosome The third and fourth chromosome have the same structure 5 loci are located on each of the two chromosomes and the distance between adjacent loci is 20 cM 2 3 6 Temporal parameters An important feature of quantiNEMO is that some parameters may change over time during a simulation Such parameters are indicated as temporal in this manual Temporal arguments are enclosed within two parentheses For each change of CHAPTER 2 USING QUANTINEMO 11 the argument over time a pair consisting of a generation index and a corresponding argument is needed The values of a pair are separated by at least one space Pairs are separated by a coma or by a semi colon The fir
38. allows to define any simulation The format of this settings file is described in section 2 3 There are several possibilities to pass the settings file to the executable standard A simulation is launched by double clicking on the executable or by writing the name of the executable e g quantinemo exe for Windows to a console window quantiNEMO will prompt you for the name of the settings file default file If the settings file has the default name quantinemo ini and the set tings file resides in the folder of the executable quantiNEMO will automatically use this settings file to perform the simulation when quantiNEMO is launched CHAPTER 2 USING QUANTINEMO 6 parameter The settings file name may be passed as a parameter to the executable when quantiNEMO is launched This can be done using a console window gt quantinemo exe settings ini Depending on the operating system a in front of the executable is some times needed to specify that the executable is located in the current directory Linux gt quantinemo settings ini 2 3 Input the settings file This section describes the format of the settings file The settings file is a text file with in general one parameter per line in a key value scheme For example patch_capacity 1000 sets the parameter patch capacity to the value of 1000 The order of appearance of the parameters In the settings file does not matter However a
39. anti_environmental_model set to 2 or 3 see also parameter quanti_heritability the additive genetic variance may also be directly output as summary statistic stat option q var A or indirectly in the population differentiation measurement Qsr stat option q qst The additive genetic variance Va is computed following Lynch and Walsh 1998 p85 87 For a quantitative trait determined by a single locus this is nbAllele Vi 5 Dia i 1 Where p is the allele frequency of allele a is the additive effect of allele and a is the average excess of allele Note that we show here the computation for a single locus for simplicity reasons In quantiNEMO a full version for traits determined by multiple loci is implemented However the implementation does not take into account linkage between loci thus the additive genetic variance is underestimated when loci are linked O for any case If this model is chosen the additive effect a and the av erage excess a are computed While the average excess is simple to compute the computation of the additive effect requires a time consum ing least square regression This formula is valid for any mating system but its computation is rather slow Note that the least square regression has not always a solution thus the additive effects a are not always computable In this case the locus in question is skipped from the analysis and a warning is returned WARNING Va cou
40. apacity patch_capacity_fem patch_capacity_mal integer matrix temporal These parameters allow to set the carrying capacities of the patches The car rying capacity of a patch is the maximal number of individuals that a patch may support The carrying capacities may vary among sexes patches and time The carrying capacities have to be specified either for each sex sep arately parameters patch capacity fem and patch capacity mal or for both sexes together parameter patch capacity In the first case both sex specific parameters have to be set if two sexes are simulated In the latter case the carrying capacities for females and males are assumed to be identical i e the carrying capacity of females and males is patch_capacity 2 If hermaphrodites are simulated the parameters patch_capacity and patch_capacity_fem are iden tical In case all three parameters are set only the sex specific parameters will be used as they are more informative To set the carrying capacities in dividually for each patch a matrix is needed The matrix is adjusted to the number of patches if necessary see section 2 3 5 4 If all patches have the same carrying capacities a single number as argument is sufficient to specify the 38 CHAPTER 5 METAPOPULATION 39 carrying capacities If the initial population sizes parameter patch_ini_size are not specified the simulation will start with all the populations set at carrying capacity Examples The following
41. ary statistic name in the output file where T is the index of the trait The summary statistic name column Stat name may be used to specify the sum mary statistic to be computed e g q varA Similar summary statistics within a thematic group may be obtained at once using the name within square brackets after the group title e g quanti Using this group statistic name all summary statistics of the thematic group marked with a star will be computed Table 6 1 Summary statistics available for quantitative traits Stat name Description Quantitative trait statistics quanti available only for adults q VgW genetic variance within patches q VgB genetic variance between patches q VpW phenotypic variance within patches q VpB phenotypic variance between patches q VaW additive genetic variance within patches q qst Qsr g varA p additive genetic variance of patch following Lynch and Walsh 1998 p85 87 computed for each patch q meanG_p genetic mean of patch i computed for each patch q varG_p genetic variance of patch i computed for each patch q meanP p phenotypic mean of patch computed for each patch d varP p phenotypic variance of patch 7 computed for each patch Genotype coancestry q adlt off coa q adlt off theta mean within patch coancestry q adlt off alpha mean between patch coancestry q adit off thetaFF mean within patch within females coancestry Table 6 1 continued on next page
42. assed as unique parameter to the script quanti geno value sex 0 1 2 default 0 This parameter allows to choose which sex is output 0 Both Output includes both sexes 1 Females Output includes only female genotypic values 2 Males Output includes only male genotypic values quanti geno value age 0 1 2 default 0 This parameter allows to choose which age is output 0 Adults Output includes only adult genotypic values 1 Juveniles Output includes only juvenile genotypic values 2 Both Output includes juveniles and adults genotypic values 6 7 3 Phenotypic value Similar to the genotypes the phenotypic values of the adults may be periodically dumped to files The phenotype of juveniles cannot be output as the phenotype is only computed when selection acts and this is at the reproduction stage i e when individuals are adults The output files will be stored in the folder given by the parameter quanti_phenot_dir and will have the name of the base file name see parameter filename in section 4 The extension is phe A counter for the generation e g g05 and the replicate e g _r4 is inserted before the extension An example of such a file name is simulation g05r4 phe quanti_save_phenotype 0 1 2 default 0 This parameter specifies the output of the phenotype 0 None No output is generated 1 FSTAT The output contains the phenotypes in the FSTAT like format Goudet 1995 2 FSTAT ex
43. assigned patch i e this is also the proportion of emigrants of a patch which migrate to the same non natal patch A probability of 1 CHAPTER 3 LIFE CYCLE 31 means that all emigrants migrate to the same non natal patch while a value of O means that all emigrants migrate to any patch but the natal and the propagule patch Example The following example defines a metapopulation of 12 patches arranged in a two dimensional grid shown below patch_number 12 dispersal_model 3 dispersal_lattice_dims 4 3 dispersal_rate 0 1 3 6 1 Density dependent dispersal rate By default the dispersal rate is not influenced by the population size The following parameters allow to define a generalized logistic function Richards 1959 to specify a relationship between the dispersal rate and the population density population size divided by the carrying capacity The generalized logistic function defines a factor f which is then multiplied with the dispersal rate parameter dispersal_rate max min 1 S x e DriasD 1 s f min Where min is the lower asymptote parameter dispersal min max is the upper asymptote parameter dispersal_max r is the growth rate parameter dispersal_growth_rate Di is the population density with the maximal slope parameter dispersal max growth D is the current population density N K and s defines the symmetry of the curve parameter dispersal_symmetry CHAPTER 3 LIFE CYCLE 32
44. bility to mutate to this allele 1 SSM Single Step Mutation In contrast to the K Allele Model the new allele depends on the current allele When a mutation occurs the current allele is replaced by one of its neighboring alleles concerning the allele index For example if the allele with the index 12 mutates it changes either to the allele with the index 13 or to the allele with the index 11 The boundaries are reflexive i e the allele index can not exceed the range of alleles i e 1 ntrl_all In line with the Increment Mutation Model IMM for quantitative traits it is possible to specify mutation probabilities of the steps to mutate using the allelic file see section 7 1 In this case the number of alleles ntrl_all has to be odd as for the IMM Then similar to the IMM the probabilities of the step to mutate has to be defined using the allelic file where the step size is measured from the middle allele concerning its index i e ntrl all 2 The middle allele with index ntrl_all 2 has to have a mutation probability of 0 as a mutation of step zero is not a mutation This allows generating any mutation pattern as for instance the Generalized stepwise mutation model Estoup et al 2002 ntrl_mutation_rate decimal matrix default 0 This parameter specifies the mutation rate per locus and generation If the CHAPTER 7 NEUTRAL MARKERS 80 argument is a single value the mutation rate for all loci is the same By passin
45. ble Depending on the number of patches this may lead to the computation of a large number of summary statistics The names of such summary statistics are marked in the output file with the index of the patch _pX respectively _pX Y for pairwise summary statistics X and Y stand for the patch index starting with 1 The summary statistics are computed for every quantitative trait If several quanti tative traits are simulated the postfix _tY is added to the summary statistic name in the output file The summary statistic name column Stat name may be used to specify the sum mary statistic to be computed e g adlt nblnd Similar summary statistics within a thematic group may be obtained at once using the name within square brackets after the group title e g adlt demo Using this group statistic name all summary statistics of the thematic group marked with a star will be computed CHAPTER 5 METAPOPULATION 47 Table 5 1 Summary statistics available for the demographic structure Stat name Description Demography adlt off demo adlt off nbInd total number of individuals in the metapopulation adit off nbFem total number of females in the metapopulation adit off nbMal total number of males in the metapopulation adit off meanInd mean number of individuals per inhabited patch adlt off meanFem mean number of females per inhabited patch adlt off meanMal mean number of males p
46. ch offspring two hermaphrodite parents are randomly assigned With probability 1 N these two hermaphrodites are identical which leads to selfing Females are used to simulate hermaphrodites 1 selfing hermaphrodite For each offspring a hermaphrodite is ran domly assigned to self fertilize The parameter mating_proportion allows to set the proportion of outcrosses quantiNEMO controls that the pro portion of outcrosses is met i e that outcrossing does not result by chance 1 N in selfing Females are used to simulate hermaphrodites 2 random mating promiscuity This is random mating with two sexes For each offspring a father and a mother are randomly assigned 3 polygyny Depending on the parameter mating_males only one default or several males per patch may reproduce This fixed number of reproduc tive males are selected randomly depending on their fitnesses i e the re productive males have on average a higher fitness Then for each offspring a mother and one of these reproductive males are randomly assigned de pending on their fitnesses Thus reproductive males with higher fitnesses have a higher reproductive success among the reproductive males If no selection acts parameter breed_model set to 3 the reproductive males are randomly chosen all males have the same probability and each male has the same probability to father an offspring The parameter mat ing_proportion allows to set the proportion of random matings betwee
47. e Different types of neutral markers and or quantitative trait loci can be combined within the same simulation Genetic map quantiNEMO has an underlying genetic map which may consist of several chromo somes This allows an explicit positioning of all types of loci on the map quantitative trait loci QTL and neutral markers Metapopulation quantiN EMO allows simulating realistic population dynamics Population sizes may vary in space and in time The user can choose between several preset migration models island 1 D stepping stone 2 D stepping stone or specify the full migra tion matrix The migration pattern can change over time allowing to investigate scenarios of population fragmentation Lifecycle Each individual undergoes a single life cycle non overlapping generations The life cycle is fixed in contrast to NEMO Guillaume and Rougemont 2006 and starts with breeding and reproduction Several mating systems are available random mat ing or selfing for hermaphrodites promiscuity monogamy or polygyny for doecious gonochoric species Selection acts on the reproductive fitness of individuals After reproduction juveniles may disperse to other populations and then population size is possibly regulated Environmental stochasticity can also be introduced where populations may go extinct due to an external factor independent of population size or genetic constitution of the population Selection Several modes of se
48. e individual s info one individual per line The first number is the patch number of the individual followed by the phenotype value for each trait Each line ends with six columns consisting supplementary information on the individual see above if the parameter guanti save phenotype is set to 2 quanti phenot dir string default This parameter allows to specify the subdirectory where phenotypes are stored This directory has to be specified relative to the simulation folder parameter folder and may also contain subdirectories If not specified default the output is stored in the simulation folder parameter folder quanti_phenot_logtime integer temporal default 1 This parameter specifies the time interval of the phenotype output Since the parameter may change over time the output may be generated at any generation quanti_phenot_script string default It is possible to launch a script just after the phenotype file is generated The argument of the parameter is the file name of the script The name of the phenotype file is passed as unique parameter to the script quanti phenot sex 0 1 2 default 0 This parameter allows to choose which sex is output O Both Output includes both sexes 1 Females Output includes only female phenotypes 2 Males Output includes only male phenotypes CHAPTER 6 QUANTITATIVE TRAITS 72 quanti_phenot_age 0 1 2 default 0 This parameter allows to choose which
49. e rules as in the settings file except that line breaks are ignored i e the character between lines is not necessary Here is an example settings file disp_rate dispersal_file txt external file named dispersal_file txt dispersal rates 0 2 0 0 0 0 6 O AAN eo gt ANO OO BE D OO S eo oo oS Noor HE DW WA He CHAPTER 2 USING QUANTINEMO 12 2 4 Output files 2 4 1 Types of output quantiNEMO can generate the following types of outputs summary statistics quantiNEMO provides summary statistics for the different simulation components including genetic variance estimates quantitative trait analysis e g h Va genetic diversity and F statistics for all types of loci neutral and QTL Most of the summary statistics are available for juveniles and adults The summary statistics can be computed for any generation during the simulation The summary statistics can be saved in two ways Either they are stored for each replicate separately and or the summary statistics are averaged across replicates raw data quantiNEMO can also produce files with the raw genetic and phenotypic data The genotypes at all loci can be dumped to file in the FSTAT format Goudet 1995 Phenotypes as well as the additive dominance and epistatic effect values can be written to a file and then analyzed with any population or quantitative genetic software e g to get
50. ecify a parameter for the fifth type one has to add a 5 to the parameter name In contrast if for the fifth trait no parameter with he suffix 5 is passed quantiNEMO checks if the parameter is passed for the fourth type suffix _4 If this is also not the case quantiNEMO checks if the parameter is passed for the third type suffix _3 and so fort until a parameter is found Note that a parameter without a suffix is the same as the parameter with the suffix _1 CHAPTER 7 NEUTRAL MARKERS 82 This behavior of quantiNEMO allows to specify parameters for a group of types of neutral markers An example may make it more clear ntrl_nb_trait 2 SNP type 1 ntrl_loci_1 100 ntrl_all_1 2 STR type 2 ntrl_loci_2 10 ntrl_all_2 20 same mutation rate for SNPs and STRs ntrl_mutation_rate 0 0001 In this example we simulate 2 types of neutral markers SNPs and STRs microsatel lites We simulate 100 SNPs each with two alleles and 10 STRs each with up to 20 alleles The mutation rate is assumed to be the same for both types of markers i e 0 0001 7 5 Genetic map quantiNEMO has an underlying genetic map which may consist of several chromo somes This allows to explicitly position all types of loci on the map quantitative trait loci QTL and neutral markers The unit of the genetic map is centi Morgans Haldane 1919 cM If the genetic map is not specified by either of the following parameters
51. ed on three out of four chromo somes The loci are randomly positioned on the first 100 cMs of each chro mosome No loci of this type of neutral marker are located on chromosome 2 For every simulated replicate where a genetic map is simulated the entire genetic map including QTLs and neutral loci is dumped to a file named genetic_map txt stored in the simulation folder If multiple replicates are simulated a replicate counter example _r3 is inserted before the extension Using the parameter ge netic map output it is possible to suppress this output 7 6 Genotype output The neutral genotype of all or a part of the individuals may periodically be stored to files similar to the genotype of quantitative traits The output files will be stored in the folder given by the parameter ntrl_genot_dir and will have the name of the base file name see parameter filename in section 4 The extension is dat A counter for the generation e g _g05 and the replicate e g 14 is inserted before the extension An example of such a file name is simulation_g05_r4 dat ntrl_save_genotype 0 1 2 default 0 This parameter specifies the output of the neutral genotype 0 None No output is generated 1 FSTAT The output contains the genotypes in the FSTAT format Goudet 1995 2 FSTAT extended Same as point 1 but the file contain the following six additional columns the age class 1 offspring 2 adult the sex 0 male
52. er inhabited patch adlt off sexRatio sex ratio Gane adlt off nbPops number of inhabited patches adlt off nbInd_p number of individuals in patch i adlt off nbFem_p number of females in patch i adlt off nbMal_p number of males in patch Patch extinction ext rate ext rate proportion of extinct patches in the metapopulation Fecundity fecundity available only for adults fem meanFec mean realized female fecundity fem varFec mean variance of realized female fecundity mal meanFec mean realized male fecundity mal varFec mean variance of realized male fecundity Kinship adlt off kinship adlt off fsib mean proportion of full sib adlt off phsib mean proportion of paternal half sib adlt off mhsib mean proportion of maternal half sib adlt off nsib mean proportion of non sib adlt off self mean proportion of selfed offspring Migration migration available only for adults emigrants mean number of emigrants per patch immigrants mean number of immigrants per patch residents mean number of residents per patch immigrate mean effective immigration rate per patch ET i colonisers mean number of colonizers per extinct patch colon rate mean effective colonization rate of extinct patches Fitness fitness available only for adults VwW variance of the fitness of adults within patches Table 5 1 continued on next page CHAPTER 5 METAPOPULATION 48 Stat name Description VwB variance of the fitn
53. ess of adults between patches mean W p mean fitness of adults in patch i computed for each patch var W_p variance of the fitness of adults in patch i computed for each patch Table 5 1 Summary statistics available for the demographic structure continued Chapter 6 Quantitative traits quantiNEMO allows the simulation of multiple quantitative traits each having its own specifications Each quantitative trait is defined by one to many loci each with up to 256 alleles The allelic effects at each locus can be drawn from a normal distribution or can be set explicitly Mutations are implemented with several models The trait determinism can be purely additive or include dominance and or epistatic interactions among loci Environmental effects can also be set in different ways P G E nbLoci Gi129 441 EA ar 5 ai ay ki lav a l i 1 Where P is the phenotype of the trait G the genotypic value and the envi ronmental contribution to the trait Gim is the genotypic value of genotype 11 22 i4 1 and I are the alleles of locus 1 2 and 2 are the alleles of locus 2 a is the effect of the first allele of locus i ay is the effect of the second allele of locus 1 kr is the dominance value between allele i and 7 and 11129 is the 44 epistatic value of genotype 11 22 27 Each quantitative trait can have its own selection pressure which may be stabilizing or directional selec
54. eter mating_proportion to 0 9 These settings will lead to 90 selfing and 10 random mating Note that quantiNEMO controls that the ratio is met i e that selfing does not occur by chance probability would be 1 N when random mating should occur mating males integer default 1 This parameter sets the number of males that will be available for mating within each patch The parameter will only be used if the mating system is polygyny parameter mating system set to 3 The range of values is between 1 a single male mates and the carrying capacity of the males all males may mate sex ratio decimal default 1 This parameter allows to specify the ratio of males to females of the offspring in a patch If hermaphrodites are simulated parameter mating_system is set to 0 or 1 the sex ratio is not considered respectively set to 0 females are used to simulate hermaphrodites 3 2 Statistics After breeding has taken place it is possible to record summary statistics specified by the parameter stat Most of the summary statistics can be recorded for offspring and or adults and for females and or males It is also possible to set the frequency parameter log time of the recording At the end of a simulation the summary statistics are written to a text file There is the choice to print the summary statistics individually per replicate and or summed up across replicates mean and variance CHAPTER 3 LIFE CYCLE 26 across replicates In
55. eter ntrl_allelic_file and especially column keyword col_ini_freq If the allele frequencies are not set explicitly in the allelic file the initialization is performed following the parameter ntrl_ini_allele_model ntrl ini genotypes string default This parameter allows to specify a name of an FSTAT file Goudet 1995 con taining the initial genotypes of the individuals for each population If such a file is present the initialization of the metapopulation is done solely by this file ignoring the parameters ntrl_ini_allele_model and patch ini size A sin gle FSTAT file is needed for all types of neutral markers together containing the total number of loci of all neutral markers together Note that quan tiNEMO allows to output an appropriate file for any generation see param eter ntrl save genotype This allows to resume a simulation to generate tailored initial conditions or to continue a simulation with modified settings CHAPTER 7 NEUTRAL MARKERS 81 If the parameter ntrl_save_genotype is set to 2 an extended FSTAT file is generated Also this file may be used to initialize the metapopulation In this case quantiNEMO overtakes the supplement information provided by the file especially the sex and age of the individual the index of the individual its mother and father Thus the supplement information allows to resume a simulation without the loss of the pedigree Note that for an entire resume of a simulation also the gen
56. example defines a metapopulation of 5 patches each with a carrying capacity of 100 individuals patch_number 5 patch capacity 100 If the carrying capacities vary between patches the following definition with a matrix may be used patch_number 5 patch_capacity 100 200 300 400 500 In the following example however the matrix of the parameter patch_capacity will be adjusted by quantiNEMO to meet the number of patches This will result in 10 patches with the carrying capacities 100 200 300 400 500 100 200 300 400 and 500 patch_number 10 patch_capacity 100 200 300 400 500 If the carrying capacities are defined for each sex separately both sex specific parameters must be present if two sexes are simulated patch_number 5 patch_capacity_fem 100 200 300 400 500 patch_capacity_mal 200 400 600 800 1000 In the following example the carrying capacity is specified with too many parameters As the sex specific parameters are more informative they have precedence over the global setting of carrying capacities 1000 which will therefore be ignored patch_number 5 patch_capacity 1000 patch_capacity_fem 100 200 300 400 500 patch_capacity_mal 200 400 600 800 1000 patch_ini_size patch ini size fem patch ini size mal integer matrix These parameters allow to set the initial population sizes of the patches i e the populations sizes present at the beginning o
57. f the simulation These param eters are optional If none of these parameters is set i e the initial population CHAPTER 5 METAPOPULATION 40 sizes are not set the simulation will start with all the population sizes set at carrying capacity The initial population sizes may vary among sexes and patches The initial population sizes have to be specified either for each sex separately parameters patch ini size fem and patch_ini size mal or for both sexes together parameter patch_ini_size In the first case both sex specific pa rameters have to be set if two sexes are simulated In the latter case the initial population sizes for females and males are assumed to be identical i e the ini tial population size of females and males is patch ini size 2 If hermaphrodites are simulated the parameters patch_ini_size and patch_ini_size_fem are identical In case all three parameters are set the sex specific parameters will be used as they are more informative To set the initial population sizes individually for each patch a matrix is needed The matrix is adjusted to the number of patches if necessary see section 2 3 5 4 If all patches have the same initial population sizes a single number as argument is sufficient to specify the initial population sizes 5 2 Selection pressure Phenotypes for quantitative traits see section 6 may be under selection Selection pressures may vary among quantitative traits sexes patches and time To specif
58. for each patch and quantitative trait patch_dir_sel_max patch dir sel max fem patch dir sel max mal decimal matrix temporal default 1 These parameters allow to set the upper asymptote of the selection curve for each patch and guantitative trait patch dir sel growth rate patch dir sel growth rate fem patch dir sel growth rate mal decimal matrix temporal default 1 These parameters allow to set the slope of the selection curve for each patch and quantitative trait If the argument is positive larger phenotypes have a higher fitness while if negative smaller phenotypes have a higher fitness patch dir sel max growth patch dir sel max growth fem patch dir sel max growth mal decimal matrix temporal default 0 These parameters allow to set the phenotype with the maximal growth patch dir sel symmetry CHAPTER 5 METAPOPULATION 44 patch_dir_sel_symmetry_fem patch_dir_sel symmetry_mal decimal matrix temporal default 1 These parameters allow to set the symmetry of the curve The default value of 1 results in a symmetric slope patch_dir_sel_growth_rate_var decimal matrix default 0 This parameter specifies the variance of the normal distribution by which the selection slope varies at each generation e g annual fluctuations of the mean temperature By default the local selection slope does not vary patch dir sel max growth var decimal matrix default 0 This parameter specifies the
59. g a matrix of mutation rates it is possible to set the mutation rate for each single locus individually By default no mutations occur ntrl_mutation_shape integer default 0 This parameter allows to specify the shape of the gamma distribution from which the mutation rates for each locus are drawn The mean mutation rate is given by the parameter ntrl_mutation rate D shape mut_rate shape Where T is the gamma function requiring two parameters The first parameter defines the shape of the gamma distribution parameter ntrl mutation shape and the second parameter the scale to the gamma distribution The scale parameter is set such does the mean of the gamma distribution is equal to the mean mutation rate mut_rate respectively parameter ntrl_mutation_rate If the shape is 0 default value all loci have the same mutation rate defined by the parameter ntrl_mutation_rate The parameter ntrl_mutation_shape is ignored if the mutation rates parameter ntrl_mutation_rate are explicitly defined for each locus by a matrix 7 3 Initial genotpyes There are several methods to set the allele frequencies or even the genotypes of the in dividuals at the start of a simulation initialisation The genotypes of the individu als may be set using an FSTAT file Goudet 1995 parameter ntrl_ini_genotypes If such a FSTAT file is not present the genotypes are randomly drawn following the explicitly set allele frequencies in the allelic file see param
60. ge while the probability to stay at home is 1 m The parameter dispersal border model allows to specify how to treat the bor der patches and the parameter dispersal_lattice_dims allows to specify the dimensions of the grid Models and their specific parameters model additional parameters 0 Migrant pool Island 1 Propagule pool Island dispersal propagule prob 2 1D Stepping Stone dispersal border model 3 2D Stepping Stone dispersal_border_model dispersal lattice range dispersal_lattice_dims CHAPTER 3 LIFE CYCLE 30 dispersal lattice range 0 1 default 0 This parameter sets the number of neighboring patches used for dispersal The dispersal probabilities to these adjacent patches are m 4 in the first case and m 8 in the second This parameter is only used in the 2D Stepping Stone model parameter dispersal_model set to 3 0 4 neighbors 4 adjacent patches up down left and right 1 8 neighbors 8 adjacent patches as before plus the diagonals dispersal_border_model 0 1 2 default 0 This parameter specifies how the patches at a border of the Stepping Stone model should be treated 0 Circle Torus In the 1D Stepping Stone model the first and last patches are connected to each other by migration leading to a circle In the 2D Stepping Stone model individuals of an edge patch may migrate to the other side leading to a torus donut world This means that there are no edges eliminating any such effects
61. genot_age 0 1 2 default 0 This parameter allows to choose which age is output 0 Adults Output includes only adult genotypes 1 Juveniles Output includes only juvenile genotypes 2 Both Output includes juveniles and adults genotypes 7 7 Summary statistics The summary statistics listed in the table below are available for neutral markers The column Stat name contains the name of the summary statistic used to specify which summary statistics are computed parameter stat for details see section 3 2 These names appear also in the output file The column Description contains a short description of the summary statistic Some of the summary statistics are available for adults and offspring indicated by adit off To obtain a certain summary statistic for adults the prefix adlt has to be added to the summary statistic name e g adlt allnb respectively the prefix off to obtain the summary statistic for offspring e g off allnb Some of the summary statistics are computed for each patch separately indicated by computed for each patch in the table others for each pair of patches indicated by all pairwise combinations computed Depending on the number of patches this may lead to the computation of a large number of summary statistics The names of such summary statistics are marked in the output file with the index of the patch _pX respectively _pX Y for pairwise summary statistics X and Y stand for the
62. he adults are removed Only the juveniles remain in the model Regulation Before dispersal some patches may be overcrowded This regu lation stage allows to control the population sizes Dispersal Juveniles may migrate to other patches and become adults Regulation After dispersal some patches may be overcrowded This regula tion stage allows to control the population sizes Extinction Due to stochastic events populations may go extinct In the following the life cycle events and their options are described in details 20 CHAPTER 3 LIFE CYCLE 21 Y CA Breeding Regulation Dispersal re 4 ie Regulation 3 1 Breeding This stage performs mating and breeding of the new offspring generation following the mating system chosen Adults are not removed here adults are removed in the event aging This is the event were selection acts The reproduction model implemented in quantiNEMO is composed of two phases In the first phase for each patch the total number of offspring to be produced is computed This number of off spring depends on the two parameters breed_model and mating_nb_offspring_model In a second phase for each offspring to be produced a pair of local of the same patch parents is assigned This assignment of the parents to the offspring depends on the mating system parameter mating system and on the fitness of the parents if selection acts Thereby adults with a higher fitness have on ave
63. hromosomes and their lengths The loci are then randomly positioned on the chromosomes The matrix con tains the length of each chromosome in cM It is possible to skip chromosomes if they don t contain loci see section 2 3 5 4 guanti loci 11 quanti locus positions random 4 1 100 3 100 100 In this example the quantitative trait is defined by 11 loci located on three out of four chromosomes The loci of the quantitative trait are randomly po sitioned on the first 100 cMs of each chromosome No loci of this quantitative trait are located on chromosome 2 For every simulated replicate where a genetic map is simulated the entire genetic map including QTLs and neutral loci is dumped to a file named genetic_map txt stored in the simulation folder If multiple replicates are simulated a replicate counter example _r3 is inserted before the extension Using the parameter ge netic_map_output it is possible to suppress this output 6 7 Output Apart from the genotypes and the phenotypes it is also possible to obtain certain summary statistics the allelic the dominance and the epistatic effects For each type of output it is possible to define individually the time interval at which the output is generated which may also vary over time 6 7 1 Genotype The genotype of all individuals may periodically be dumped to files The output files will be stored in the folder given by the parameter quanti_genot_dir and wil
64. in brackets separated by space The available key words and their corresponding summary statistics are listed in the corresponding simulation component sec tion in this manual Summary statistics about the demography may be found in section 5 3 summary statistics about quantitative traits may be found in section 6 8 and summary statistics about neutral markers may be found in sec tion 7 7 If no arguments are specified this event will be skipped If summary statistics are specified which cannot be computed since the corresponding component is missing e g F statistics of neutral loci can only be computed if neutral markers are simulated a warning will be given stat n adlt fstat q adlt fstat quanti adlt demo In this example the summary statistics for the key words n adlt fstat quanti and adlt demo are considered by quantiNEMO while the key word q adlt fstat is commented out 3 3 Output quantiNEMO can also produce files with the raw genetic and phenotypic data Geno types of neutral markers and quantitative traits can be dumped to file in the FSTAT format Goudet 1995 For quantitative traits the phenotypes may also be writ ten to a file for any generation sex or age The selection of the various outputs is done in the corresponding simulation components neutral genotype see section 7 6 quantitative trait genotype see section 6 7 1 and quantitative trait phenotype see section 6 7 3 If no output is desired
65. ion quanti_allelic_file string default gt This parameter allows to pass the name of a file containing allelic informations such as the allelic effects the mutation frequency and or the initial frequency The information passed by this file has precedence over other settings however the number of alleles and loci has to be in line with the parameters quanti_loci and quanti_all The information can be set globally for all loci if they have the same specifications or for each locus separately The allelic file has the following format Allelic file FILE_INFO col_locus 1 col_allele 2 col_allelic_value 3 col mut freg 4 col ini freg 5 i locus allele value mut freg ini_freq 1 1 1 0 0 2 0 0 2 1 2 0 5 0 2 0 0 2 1 3 0 0 0 2 1 0 2 1 4 0 5 0 2 0 0 2 1 5 1 0 0 2 0 0 2 2 1 2 0 2 0 0 2 CHAPTER 6 QUANTITATIVE TRAITS 51 2 2 f 0 2 0 0 2 2 3 0 0 2 1 0 2 2 4 i 0 2 0 0 2 2 5 2 0 2 0 0 2 The file has to start with a file information box FILE_INFO This box con tains informations about the structure of the following table and thus allowing flexibility in the format of the table For example the order of the columns in the table may vary or some columns may be ignored The file information box starts with the key word FILE_INFO This key word is followed by brackets within which the user has to specify the contents of the columns to be considered by quantiNE
66. is example now selection acts on the reproduction stage soft selection The fitness is computed on the simulated quantitative trait The trait consists of a single locus with up to 255 alleles Allelic effects are normally distributed with a variance of 1 The phenotype of the quantitative trait is determined by pure additive effects of the alleles Stabilizing selection acts on the phenotype with an optimum of 0 and a variance of 1 A single neutral marker locus is simulated with up to 2555 alleles Both the quantitative trait locus and the neutral marker locus do not mutate for both markers some common summary statistics are computed 2 6 Simulation example This section describes a more realistic simulation example and describes how the output is stored The settings file of this example named quantiNemo_example ini is included in the compressed folders of the downloads of quantiNEMO generations 100 metapopulation patch_number 10 patch_capacity 1000 dispersal_rate 0 01 mating breed_model 0 mating system 0 selection patch_stab_sel_optima 10 patch_stab_sel_intensity 1 quantitative trait quanti loci 5 quanti_all 255 CHAPTER 2 USING QUANTINEMO 15 quanti_mutation_model 0 guanti mutation rate le 4 quanti_save_genotype 2 guanti genot logtime 10 guanti genot dir quanti genotype quanti_save_phenotype 2 quanti_phenot_logtime 10 quanti_phenot_dir quanti_phenotype neutral marker ntrl_loci 5 n
67. itative traits CIl Architecte AE hk N ee N a wee 611 Allelic effects CONTENTS 6 2 6 3 6 4 6 5 6 6 6 7 6 8 6 1 2 Dominance chet 4e las du BES A A Gls rd AR ME AA AAR AR EE ES Oe eut NE o oe be Ere De RE a Don ee eee RR Initial genotypes Multiple traits Selection models 6 7 1 Genotype 6 7 2 Genotypic A EEE Pee 41 6 7 3 Phenotypic ME RE oi otek RR GTA Architechite oo EE db due da RR EER ORE EE wee Summary statisties AA dax av ch EE EE OU 7 Neutral markers 7 1 Architecture 7 2 Mutation 7 3 Initial genotpyes 7 4 Multiple traits 7 5 Genetic map 7 6 Genotype output 7 7 Summary statistics Bibliography Index 111 92 54 56 59 62 63 64 65 66 66 68 70 72 72 76 76 78 80 81 82 83 85 88 90 Chapter 1 Introduction 1 1 Scope quantiNEMO is an individual based genetically explicit stochastic simulation pro gram It was developed to investigate the effects of selection mutation recombina tion and drift on quantitative traits with varying architectures in structured popula tions connected by migration and located in a heterogeneous habitat quantiNEMO is highly flexible at various levels population selection trait s architecture genetic map for QTL and or markers environment demography mating system etc 1 2 Availability The website http www unil ch popgen softwares quantinemo includes executables
68. ite previous outputs a counter is added to the file name between base CHAPTER 2 USING QUANTINEMO 13 name and extension There are two types of counters the generation counter and the replication counter A counter is only added if there is a risk of overwriting For example the replication counter is only added if several replications are performed The generation counter starts with _g and the replication counter with 1 These characters are followed by the number of the generation and replication respectively Note that generations and replications start at 1 The number has as many digits as are needed to represent the highest number in the simulation simulation_g0001_r01 dat simulation_g0002_r01 dat simulation_g5000_r10 dat 2 5 Minimal settings file Most of the parameters have default values This allows to make short and clear settings files This section describes the parameters needed for a minimal settings file and describes the simulated model respectively how the default values are set The following two parameters are needed in every settings file generations 500 patch_capacity 1000 This minimal settings file allows to perform a simulation with a single population consisting of 1000 hermaphrodites The population evolves under neutral random mating for 500 generations keeping the population size constant at carrying capac ity No genetic data are simulated since no quantitative
69. l In CHAPTER 3 LIFE CYCLE 23 case of soft selection breed model set to 0 this parameter specifies how to compute the total number of offspring for each patch separately In case of soft hard selection breed_model set to 1 this parameter specifies how to com pute the total number of offspring of the entire metapopulation This total number of offspring of the entire metapopulation is then distributed among the patches based on the mean fitnesses of the patches In case of hard selection breed_model set to 2 this parameter specifies the total number of offspring per patch assuming a maximal fitness of 1 for all adults 0 carrying capacity Nog K The total number of offspring Nopp is set to the carrying capacity K parameter patch capacity of the patch or the metapopulation respec tively 1 keep number Nos N The total number of offspring No rf corresponds to the number of adults N ie the number of individuals is kept constant Note that a reg ulation of the patch densities after dispersal can lead to an unwanted continuing reduction of the entire metapopulation size 2 fecundity Nosf Npf The number of offspring Nos depends on the mean fecundity of the females f defined by the parameter mean_fecundity 3 fecundity stochastic Nos Poisson N p f Same as point 2 but the computation of the total number of offspring has a stochastic component 4 logistic regulation Noss WER The total number of
70. l have the name of the base file name see parameter filename in section 4 The extension is dat A counter for the generation e g _g05 and the replicate e g 14 is inserted before the extension An example of such a file name is simulation_g05_r4 dat quanti save genotype 0 1 2 default 0 This parameter specifies the output of the quantitative genotype at the QTLs CHAPTER 6 QUANTITATIVE TRAITS 67 0 None No output is generated 1 FSTAT The output contains the genotypes in the FSTAT format Goudet 1995 2 FSTAT extended Same as point 1 but the file contain the following six additional columns the age class 1 offspring 2 adult the sex 0 male 1 female the ID of the individual the ID of the mother the ID of the father and the fitness of the individual The ID is a unique identifier for each individual of a simulation in the format 345_23 meaning that this is the 345th individual born in patch 23 The IDs of the individual the mother and the father allow to extract pedigree informations if the output is stored for each generation and also to investigate the migration behavior of the individual and its parents An example of such a file with quanti save genotype set to 2 5 4 20 2 loc 1 loc 2 loc 3 loc 4 1 1415 1019 2002 0820 1 1 10 1 1 1 01 0 345 1 0814 0219 2002 2020 1 1 11 1 8 1 2 4 0 334 1 0808 0217 1902 0820 1 1 12_1 5 3 5 1 0 123 5 1004 0917 1404 1
71. ld not be correctly estimated 1 time at generation 20 see manual parameter quanti_va_model This warning is returned the first 10 times the problem occurs and then every hundredth time It is up to the user to decide whether the prob lem occurs to often or if these missing points are acceptable If the environmental variance is set by the narrow sense heritability parameter quanti_environmental_model set to 1 or 2 and the additive genetic vari ance cannot be computed quantiNEMO uses the additive genetic variance computed using the algorithm for random matings this parameter set to 1 CHAPTER 6 QUANTITATIVE TRAITS 99 1 limited to random mating In case of random mating the additive ef fect a and the average excess a are identical Therefore the time consuming computation of the additive effect can be omitted Note that even when the mating system is set to random mating parameter mat ing system set to 0 or 2 mating is not randomly if selection acts as the choice of the parents depends on their fitnesses Its up to the user to decide whether this quick way to compute the additive genetic variance is appropriate or not 6 2 Mutation Mutation rates may be defined for each locus individually by explicitly defining the individual mutation rates parameter quanti_mutation_rate or by defining the gamma distribution from which the individual mutation rates are drawn parameters quanti_mutation_rate and
72. lection are available In stabilizing selection modes a specific optimum and selection intensity may be defined for each population and quantitative trait B rger 2000 In directional selection modes the strength and direction of selection may vary for each trait and population Furthermore the selective pressures can change over time Last selection can be soft or hard Wallace 1968 Initial settings quantiNEMO is highly flexible in the setting of the initial simulation conditions The initial population sizes may be set for each population and sex parameters patch ini size patch_ini_size_fem or patch ini size mal Furthermore the initial allele freguencies for each population may be either maximal polymorph or monomorph parameters guanti ini allele model and ntrl ini allele model or may be set explicitly for each population locus and allele parameters guanti allelic file and ntrl_allelic file CHAPTER 1 INTRODUCTION 4 1 7 Input and output Input quantiN EMO is launched using a settings file The settings file is a text file with flex ible and easy to understand structure The information is specified in a parameter argument scheme where the order of the parameters does not matter The file can be edited with any text editor and comments may be added for better readability We provide also a syntax highlighting definition for better readability Output quantiNEMO provides summary statistics for the different simulation
73. llelic file Epistatic file FILE_INFO col_genotype 1 col_epistatic_value 2 col_genotypic_value 3 genotype epistaticVal genotypicVal 0101 0101 0 242984 1 87009 0101 0102 0 580787 0 834811 5050 5048 0 264001 1 24981 5050 5049 0 118982 0 55701 5050 5050 0 071359 2 26644 The file has to start with a file information box FILE_INFO This box con tains informations about the structure of the following table and thus allowing flexibility in the format of the table For example the order of the columns in the table may vary or some columns may be ignored The file information box starts with the key word FILE_INFO This key word is followed by brackets within which the user has to specify the contents of the columns to be considered by quantiNEMO Each column is specified by a pair consisting of a key word e g col_locus followed by the column number the ordering starts with 1 Each column definition has to be on a new line The following column keywords are available col_genotype This keyword specifies the column containing the genotype This column is mandatory The genotype is enclosed by brackets and the genotype itself has to be in the FSTAT format Goudet 1995 The two alleles of a locus are written consecutively without any space For all alleles the same number of digits are needed Two different loci are separated by a space The above example consists of two l
74. mation of fixation indexes and gene diver sities Annals of Human Genetics 47 Jul 253 259 Ravigne V Olivieri I and Dieckmann U 2004 Implications of habitat choice for protected polymorphisms Evolutionary Ecology Research 6 1 125 145 Richards F 1959 A flexible growth function for empirical use J Exp Bot 10 290 300 Wallace B 1968 Polymorphism population size and genetic load in R C Lewontin ed Population biology and evolution Syracuse University Press Syracuse N Y pp 87 108 Wallace B 1975 Hard and soft selection revisited Evolution 29 465 473 88 BIBLIOGRAPHY 89 Weir B S and Cockerham C C 1984 Estimating f statistics for the analysis of population structure Evolution 38 1358 1370 Index breed_model 21 dispersal_border_model 30 dispersal_k_growth_rate 32 dispersal_k_max 32 dispersal_k_max_growth 32 dispersal_k_min 32 dispersal_k_symmetry 33 dispersal lattice dims 30 dispersal lattice range 30 dispersal model 29 dispersal propagule prob 30 dispersal rate 28 dispersal rate fem 28 dispersal rate mal 28 extinction rate 33 filename 35 folder 34 generations 34 genetic_map_output 36 growth_rate 23 logfile 35 logfile_type 35 mating males 25 mating nb offspring model 22 mating proportion 25 mating system 24 mean fecundity 24 ntrl all 76 ntrl_allelic file 76 ntrl genot_age 85 n
75. mulation 6 8 Summary statistics The summary statistics listed in the table below are available for quantitative traits The column Stat name contains the name of the summary statistic used to specify which summary statistics are computed parameter stat for details see section 3 2 These names appear also in the output file The column Description contains a short description of the summary statistic Some of the summary statistics are available for adults and offspring indicated by adit off To obtain a certain summary statistic for adults the prefix adlt has to CHAPTER 6 QUANTITATIVE TRAITS 73 be added to the summary statistic name e g adlt allnb respectively the prefix off to obtain the summary statistic for offspring e g off allnb Some of the summary statistics are computed for each patch separately indicated by computed for each patch in the table others for each pair of patches indicated by all pairwise combinations computed Depending on the number of patches this may lead to the computation of a large number of summary statistics The names of such summary statistics are marked in the output file with the index of the patch _pX respectively _pX Y for pairwise summary statistics where X and Y are the index of the patches starting with 1 The summary statistics are computed for every quantitative trait If several quanti tative traits are simulated the postfix t T is added to the summ
76. n Repetition With the command rep value number it is possible to specify a repetition of a single data point rep works similar to rep in the statistical package R The macro needs two arguments separated by a comma the first one is the data point to be repeated and the second argument specifies the number of repetitions Example patch_ini size 1000 rep 0 9 patch_ini size 1000 0 0 0 0 0000 0 Both specifications of the initial population sizes are identical i e the macro rep translates its arguments to the lower specification In this example only the first patch is populated at the start of the simulation allowing to simulate a colonization scenario 2 3 5 Parameter types There are different types of arguments that a parameter may take In the following the argument type is specified within square brackets Example stat_log_time integer 2 3 5 1 Integer Integers are whole numbers i e a dot less number The following forms are equiv alent 1000 or 1e3 2 3 5 2 Decimal Decimals are floating point numbers The following forms are equivalent 0 0001 0001 or le d 2 3 5 3 String Strings are text arguments If the string contains spaces the argument has to be enclosed within quotation marks When a string is enclosed by quotation marks it may be written over several lines Example of a string with a space folder first simulation CHAPTER 2 USING QUANT
77. n any male and female i e also males of the non reproductive group may get the chance to reproduce 4 monogamy For each female a male is randomly assigned to be its partner for all offspring If there are less females than males present in the patch not all males will mate In contrast if there are more females than males present in the patch males may belong to several mating pairs For each offspring a parent pair is randomly assigned depending on the fitness of the female if no selection acts the parent pairs have the same probability CHAPTER 3 LIFE CYCLE 25 to be selected Thus parent pairs where the females have a higher fitness have on average a higher reproductive success The parameter mating_proportion allows to set the proportion of random matings Models and their specific parameters model additional parameters 0 random mating herma 1 selfing herma mating_proportion 2 random mating prom sex ratio 3 polygyny sex ratio mating_proportion mating males 4 monogamy sex_ratio mating_proportion mating proportion decimal default 1 This parameter allows to specify the ratio of a special mating system in relation to random mating A value of 1 default means that only the special mating occurs and a value of 0 means that only random mating occurs For example if we want to simulate a plant with a selfing rate of 90 we have to set the parameter mating system to 1 selfing and this param
78. n locus In this example we are using two digit per allele the first two digits of a locus genotype number are the index of the first allele e g allele 14 for the first allele of the first locus of the first individual while the two next digits are the index of the second allele e g allele 15 for the second allele of the first locus of the first individual Each line ends with six columns consisting supplementary information on the individual see above if the parameter ntrl_save_genotype is set to 2 ntrl genot dir string default This parameter allows to specify the subdirectory where the genotypes are stored If not specified default the output is stored in the simulation folder parameter folder ntrl_genot_logtime integer temporal default 1 This parameter specifies the interval of the genotype output Since the param eter may change over time the output may be generated at any generation ntrl_genot_script string default gt It is possible to launch a script just after the genotype file is generated The argument of the parameter is the file name of the script The name of the genotype file is passed as unique parameter to the script CHAPTER 7 NEUTRAL MARKERS 85 ntrli genot sex 0 1 2 default 0 This parameter allows to choose which sex is output O Both Output includes both sexes 1 Females Output includes only female genotypes 2 Males Output includes only male genotypes ntrl_
79. neration see parameter quanti save genotype This allows to resume a simulation to generate tai lored initial conditions or to continue a simulation with modified settings If the parameter quanti_save_genotype is set to 2 an extended FSTAT file is generated Also this file may be used to initialize the metapopulation In this case quantiNEMO overtakes the supplement information provided by the file especially the sex and age of the individual the index of the individual its mother and father Thus the supplement information allows to resume a simulation without the loss of the pedigree Note that for an entire resume CHAPTER 6 QUANTITATIVE TRAITS 63 of a simulation also the genotypes of the neutral markers have to be set see parameter ntrl_save_genotype In this case the number of individuals per patch and the individuals supplement information sex age individuals id mothers id fathers id must be in agreement with each other quanti_ini_allele model 0 1 default 0 If the genotypes or allele frequencies are not already defined in another way the initialization of the genotypes may be either polymorphic where the prob ability of each allele is identically or monomorphic where all populations are fixed for a single allele 0 polymorph The populations are maximally polymorph in respect to al lele frequencies at the start of a simulation 1 monomorph The populations are monomorph in respect to allele fre quencies at
80. ng equation is used to compute the genotype nbLoci Givo2 ur Era 5 Gi i 1 Where G11 22 47 is the genotypic value of genotype 11 22 i2 1 and 1 are the alleles of locus 1 2 and 2 the alleles of locus 2 G is the genotypic effect of locus i additive and dominance effects and 11129 7 is the epistatic effect of genotype 11 22 62 unique for each multilocus genotype There are two possibilities to define these epistatic effects Either they are defined explicitly for each genotype using a separate file see parameter quanti_epistatic file or by defining the variance of the normal distribution from where the epistatic effects are randomly drawn see parameter quanti_epistatic_var Using the parameter quanti_epistatic_file it is also possible to define the genotypic effects directly If the epistatic effects are defined in both ways the explicitly defined effects are used Note that all effects have to be defined in the same way i e explicitly or by their distribution By default the parameter quanti_epistatic_var is set to zero resulting in simulations without any epistatic effects quanti_epistatic_file string default This parameter allows to pass the name of a file containing the epistatic ef fects The number of defined genotypes has to be in line with the parameters CHAPTER 6 QUANTITATIVE TRAITS 59 quanti_loci and quanti_all The epistatic file has a similar format as the a
81. ntial parameter is not the same as a temporal parameter see section 2 3 6 patch capacity 5 10 20 In this example patch capacity is a sequential parameter with three arguments If patch capacity is the only sequential parameter three consecutive simulations will be launched with identical parameter arguments except for the parameter patch_capacity which will be set to 5 for the first simulation to 10 for the sec ond simulation and to 20 for the third simulation If several parameters are sequential parameters all combinations of the sequential parameters will be simulated Example patch_number 10 50 patch capacity 5 10 20 There are two sequential parameters in this example This will result in 6 consecutive simulations with the following parameters patch_number patch_capacity 1 simulation 10 5 2 simulation 10 10 3 simulation 10 20 4 simulation 50 5 5 simulation 50 10 6 simulation 50 20 To prevent the output from overwriting the preceding simulation a unique base file name is given to each simulation This unique base file name consists of the the base file name parameter filename plus a suffix which includes the rank of the simulation If in the example above the parameter filename was set to mysim the base name for each simulation would be as follows CHAPTER 2 USING QUANTINEMO 19 filename 1 simulation mysim 1 2 simulation mysim 2 3 simulation mysim 3
82. oci each with 50 alleles Each allele is written using two digits The table must include all possible genotypes consequently the table has _nb_all x _nb_all 1 2 e rows which can be a very large number if there are many loci col epistatic value This keyword specifies the column containing the epistatic effects If this column is set the epistatic effect is added to the genotypic effect computed as described above col_genotypic_value This keyword specifies the column containing directly the genotypic effects If this column is set the genotypic effect of each genotype is set directly without taking into account allelic dominance CHAPTER 6 QUANTITATIVE TRAITS 56 and or epistatic effects If both keywords col_epistatic_value and col_genotypic_value are specified in the information box only the column col_genotypic_value is considered Note that as in all input files for quantiNEMO it is possible to add comments also in the file information box using the hash character or any text quanti_epistatic_var decimal default 0 This parameter allows to specify the variance of the normal distribution from where the epistatic effects are drawn randomly The normal distribution is centered around 0 This parameter is only taken into account if the epistatic effects are not defined explicitly by the epistatic file 6 1 4 Environment The environment may also contribute to the phenotype There are seve
83. omputed F statistics following Weir and Cockerham 1984 In adlt off fstat wc n adlt off fst we global Fsr n adlt off fis we global Fys5 n adlt off fit we global Frr n adlt off fst wc pair pairwise Fsr between patch i and j all pairwise combina tions computed Table 7 1 Summary statistics available for neutral markers continued Bibliography Beverton R J H and Holt S J 1957 On the dynamics of exploited fish popu lations Technical report U K Ministry of Agriculture and Fisherie Birger R 2000 The Mathematical Theory of Selection Recombination and Mu tation Wiley Cichester UK Estoup A Jarne P and Cornuet J M 2002 Homoplasy and mutation model at microsatellite loci and their consequences for population genetics analysis Molecular Ecology 11 9 1591 1604 Goudet J 1995 Fstat version 1 2 A computer program to calculate f statistics Journal of Heredity 86 6 485 486 Guillaume F and Rougemont J 2006 Nemo an evolutionary and population genetics programming framework Bioinformatics 22 20 2556 2557 Haldane J B S 1919 The combination of linkage values and the calculation of distances between loci of linked factors journal of genetics Journal of Genetics 8 299 309 Lynch M and Walsh B 1998 Genetics and analysis of quantitative traits Pub lisher Sinauer Associates Inc Nei M and Chesser R K 1983 Esti
84. on rate 0 0001 In this example we simulate 12 quantitative traits Traits 1 to 3 consist of 5 loci with up to 10 alleles traits 4 to 6 consist of 5 loci with up to 20 alleles traits 7 to 9 consist of 10 loci with up to 10 alleles and traits 10 to 12 consist of 10 loci with up to 20 alleles All traits have the same mutation rate of 0 0001 6 5 Selection models In the current version only selection at the reproductive stage is implemented However it is possible to modify the code in a way that selection may also act on other life cycles such as during migration or regulation The fitness of an individ ual defines its reproductive success The higher its fitness the more offspring are produced The fitness itself depends on the selection pressure and on the phenotype of the individual The better an individual is adapted to the local environment the higher its fitness is The phenotype of an individual is characterized by one to sev eral quantitative traits see section 6 4 Each quantitative trait has its individual architecture and may be selected for a different selection pressure Selection types may differ between quantitative traits The selection pressure has to be defined in the patch component see section 5 If several quantitative traits are simulated the fitness of an individual is the product of the quantitative trait s fitnesses traits W II Wi t Where W is the total fitness of the individual and W is the fitne
85. ontain subdirectories If not specified default the output is stored in the simulation folder parameter folder quanti genot logtime integer temporal default 1 This parameter specifies the time interval of the genotype output Since the parameter may change over time the output may be generated at any genera tion quanti_genot_script string default It is possible to launch a script just after the genotype file is generated The argument of the parameter is the file name of the script The name of the genotype file is passed as unique parameter to the script quanti genot sex 0 1 2 default 0 This parameter allows to choose which sex is output 0 Both Output includes both sexes 1 Females Output includes only female genotypes 2 Males Output includes only male genotypes quanti_genot_age 0 1 2 default 0 This parameter allows to choose which age is output 0 Adults Output includes only adult genotypes 1 Juveniles Output includes only juvenile genotypes 2 Both Output includes juveniles and adults genotypes 6 7 2 Genotypic value Similar to the genotypes the genotypic values may be periodically dumped to files The output files will be stored in the folder given by the parameter quanti_geno_value_dir and will have the name of the base file name see parameter filename in section 4 The extension is gen A counter for the generation e g _g05 and the repli cate e g rd is inserted
86. otypes of the quantitative traits have to be set see parameter quanti save genotype In this case the number of individuals per patch and the individuals supplement information sex age individuals id mothers id fathers id must be in agreement with each other ntrl_ini_allele model 0 1 default 0 If the genotypes or allele frequencies are not already defined in another way the initialization of the genotypes may be either polymorphic where the prob ability of each allele is identically or monomorphic where all populations are fixed for a single allele 0 polymorph The populations are maximally polymorph in respect to al lele frequencies at the start of a simulation 1 monomorph The populations are monomorph in respect to allele fre quencies at the start of a simulation All individuals are fixed for a single allele which is the middle allele i e the allele with the index ntrl_all 2 7 4 Multiple traits Different types of neutral markers can be combined within the same simulation Each type may have its own parameterization ntrl_nb_trait integer default 1 This parameter defines the number of different types of neutral markers Each type of neutral markers may have its own specifications but it is also possible to specify parameters for some types of neutral markers together If several types of neutral markers are used it is possible to address a certain type by the number of the type For instance to sp
87. particular parameter should appear only once in the settings file If a parameter appears several times in the file only the last instance is considered 2 3 1 Default value Most of the parameters have default values The default value of a parameter is taken into account when either the parameter is not listed in the settings file or its argument is missing The default values are listed behind the parameter name in this manual and are specified by default The default values are the most common setting of the parameter the default values allow to keep the settings file short and clear 2 3 2 Comments quantiN EMO allows to add comments in the settings file and all other input files There are two different types of comments Single line comments A simple hash character defines the start of a single line comment The hash character and the remaining text of the line are ignored by quantiNEMO CHAPTER 2 USING QUANTINEMO I Block comments Block comments may start and end at any place in the file This allows to comment out multiple lines at once or only a part of a line A block comment starts with the characters and ends with the characters The starting and ending characters and the text between them are ignored 2 3 3 Line end In general a parameter the key and its argument has to be written on a single line except for matrices and temporal parameters However using a backslash it is po
88. patch index starting with 1 The summary statistics are computed for every type of neutral marker If several quantitative traits are simulated the postfix _tY is added to the summary statistic name in the output file The summary statistic name column Stat name may be used to specify the sum mary statistic to be computed e g adlt fst Similar summary statistics within a thematic group may be obtained at once using the name within square brackets after the group title e g adlt fstat Using this group statistic name all summary statistics of the thematic group marked with a star will be computed CHAPTER 7 NEUTRAL MARKERS 86 Table 7 1 Summary statistics available for neutral markers Stat name Description Genotype coancestry n adlt off coal n adlt off theta n adlt off alpha n adlt off thetaFF n adlt off thetaMM n adlt off thetaFM n adlt off coa fsib n adlt off coa phsib n adlt off coa mhsib n adlt off coa nsib n adlt off theta_p n adlt off alpha pair mean within patch coancestry mean between patch coancestry mean within patch within females coancestry mean within patch within males coancestry mean within patch between sexes coancestry mean coancestry within full siblings mean coancestry within paternal half siblings mean coancestry within maternal half siblings mean coancestry within non siblings mean coancestry within patch computed for each pa
89. patterns of differentiation study linkage disequilibrium or scan for QTL Genotypes and phenotypes can be saved for any generation during the simulation log files quantiNEMO also generates log files allowing to reconstruct performed simulations There are two types of log files The first log file records the simulations performed with quantiNEMO and stores some general information This log file is stored in the folder of the executable and allows to reconstruct the chronology of performed simulations and their main features The other log file contains the used parameters is generated for each simulation separately and is stored in the simulation folder This file is in principle a copy of the used settings file and contains the starting time and the duration time of the simulation It contains also the seed see parameter seed used to initialize the simulation This file can be used as settings file to exactly repeat the performed simulation Note that due to the seed in the file the random generator will be initialized in the same way leading to the exact same values in the output 2 4 2 Naming convention All files of a simulation are stored in a unique folder see parameter folder This simulation folder may contain a substructure The name of the output files are based on the base name given by the parameter filename Depending on the type of output different extensions are added to the base name To avoid that recurring outputs overwr
90. pring of the entire metapopulation de pends purely on the parameter mating_nb_offspring model Parents are then allocated to the offspring based on their fitnesses In other words the fitness of the adults is relative to the mean fitness of the metapopu lation 2 hard selection The maximal number of offspring of a patch i e if all individuals have a fitness of 1 is defined by the parameter mat ing nb offspring model The effective number of offspring produced in a patch is however adjusted by the mean fitness of the adults of this patch ef fective number maximal_number mean_fitness The parents for each offspring are randomly drawn taking into account their fitnesses within a patch Thus fitness is directly translated into a number of offspring An example illustrating the distribution of the offspring among two popula tions depending on the chosen breeding model parameter breed_model Both populations have a carrying capacity K of 1000 individuals the parameter mating_nb_model is set to 0 and the mean fitnesses of the populations W are 0 8 and 0 4 respectively population 1 population 2 K 1000 K 1000 breeding model W 0 6 W 0 2 total 0 soft 1000 1000 2000 1 soft hard 1500 500 2000 2 hard 600 200 800 mating nb offspring model 0 1 2 3 4 5 default 0 This parameter specifies how the total number of offspring Nos is deter mined This parameter depends on the previous parameter breed mode
91. process for a specific allele CHAPTER 7 NEUTRAL MARKERS 79 Mutation rates may be defined for each locus individually by explicitly defining the individual mutation rates parameter ntrl_mutation rate or by defining the gamma distribution from which the individual mutation rates are drawn parameters ntrl_mutation_rate and ntrl mutation var There are two mutation models available Using the allelic file for neutral markers see section 7 1 it is possible to specify the probability to mutate to a certain allele explicitly for all alleles Depending on the mutation model the new allele is the drawn one KAM or its index distance from the allele index in the middle is added to the current allele index SSM A minimal defi nition for mutations requires the setting of a common mutation rate parameter ntrl_mutation_rate has a single value In this case all loci have the same mutation rate and the mutation model is KAM ntrl_mutation_model 0 1 default 0 This parameter allows to specify the mutation model The following mutation models are available 0 KAM K Allele Model At each mutation the existing allele is randomly exchanged by another allele within the range of alleles i e 1 ntrl_all By default the prob ability to mutate to any allele is the same However if the mutation probabilities are specified explicitly by the allelic file see section 7 1 the probability to mutate to a certain allele depends on the specified proba
92. quantiNEMO User Manual September 8 2011 authors Samuel Neuenschwander samuel neuenschwanderQunil ch J r me Goudet jerome goudet unil ch Fr d ric Hospital frederic hospital jouy inra fr Fr d ric Guillaume guillaum zoology ubc ca website http www unil ch popgen softwares quantinemo release 1 0 4 2011 Samuel Neuenschwander Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying provided also that the sections entitled Copying and GNU General Public License are included exactly as in the original and provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one Permission is granted to copy and distribute translations of this manual into another language under the above conditions for modified versions except that this permission notice may be stated in a translation approved by the Free Software Foundation Contents 1 Introduction 1 E A IEEE 1 12 Availability cra rara AA DU 1 e o AAA IN 1 LA A a pre ca rio FEE a sa OR T 2 1 5 Acknowledgments MERE ER ER a e 2 1 2 HE Mauw WER se ek ie OE GER See ee She EO Ra Sale EE a 2 ET bee GE LL La AR TE EE TEE ER OE ON 4 2 Using quantiNEMO 5 MR AO AE EE
93. quantiNEMO for models 1 to 4 resulting in an error message If the parameter quanti environmental model is set to 0 the parameter quanti_heritability is no longer the heritability but the environmental variance directly and can be set to 0 Using a matrix as argument it is possible to set the environmental variance respectively heritability for each patch separately quanti environmental proportion decimal default 1 This parameter specifies which environment affects the phenotype of the quan titative trait the natal or the current at the adult stage environment The argument specifies the relative weight of the current patch effect on the phe notype if the value is 1 default value only the environmental variance of the current patch affects the phenotype while if the value is 0 only the environ mental variance of the natal patch affects the phenotype CHAPTER 6 QUANTITATIVE TRAITS 58 quanti_va_model 0 1 default 0 This parameter specifies how the additive genetic variance Va of a quantita tive trait is computed If purely additive effects are simulated no dominance end epistatic effects the additive genetic variance is identical to the genetic variance Va and thus this parameter is not considered The additive genetic variance of a trait is used in several statistics For instance it is used to set the environmental variance of the patches if this variance is specified by the narrow sense heritability h2 parameter qu
94. rage a higher reproductive success If no selection acts i e no quantitative trait undergoes se lection all parents have a fitness of 1 thus the assignment of the parents to the offspring depends solely on the mating system The following parameters allow to parameterize this life cycle event breed_model 0 1 2 default 0 This parameter specifies if and how selection acts at the reproduction stage Selection acts on the phenotype of the quantitative traits see section 6 for CHAPTER 3 LIFE CYCLE 22 more details If no selection acts i e no quantitative trait undergoes selec tion all breeding models result in the same outcome respectively this param eter is ignored 0 soft selection Selection acts locally at the patch level i e patches do not interact Wallace 1975 The number of offspring of a patch depends solely on the parameter mating_nb_offspring model The parents for each offspring are randomly drawn taking into account their fitnesses within a patch Thus the reproductive output of the patch is independent of its mean fitness Individual s fitness is relative to the mean fitness of its patch 1 soft hard selection Selection acts at the metapopulation level i e the total number of offspring of the entire metapopulation is partitioned according to the mean fitnesses of the patches Ravigne et al 2004 Patches with a higher mean fitness produce on average more offspring However the total number of offs
95. ral possi bilities to set the contribution of the environment to the phenotype globally or for each patch separately Either the contribution of the environment to the pheno type is defined directly by the variance of the environmental effect model 0 by the narrow sense heritability h model 1 and 2 or by the broad sense heritabil ity H model 3 and 4 The heritability is later translated into a corresponding environmental variance Vz 1 h h Va h2 Va 1 H H Ve a Ve Where Va is the additive genetic variance computed following Lynch and Walsh 1998 p85 87 and Va the genetic variance quanti_environmental_model 0 1 2 3 4 default 0 This parameter specifies how the environmental variance is defined The fol lowing models are available 0 set Vr directly The variance of the environment is set directly by the parameter quanti_heritability which is in this case not the heritability but the environmental variance 1 Vg defined by the narrow sense heritability Vz constant The vari ance of the environment Vz is set at the beginning of a simulation gen eration 1 and is based on the narrow sense heritability h parameter quanti_heritability and the additive genetic variance V4 at generation 1 CHAPTER 6 QUANTITATIVE TRAITS 97 Note that in this case the environmental variance remains constant over time but not the heritability 2 Vg defined by the narrow sense heritability h con
96. required quantitative trait However the following parameterization is equivalent to the upper one patch_number 3 quanti_nb_trait 5 guanti selection model 0 guanti selection model 3 1 patch stab sel optima 111 2 3 2 3 4119 9 9 9 9 9 9 9 9 patch_dir_sel_growth_rate 119 9 9119 9 9 4 5 6MT7 8 7 2 3 2 In the example above the rows consisting of the number nine are read but then not taken into account since the rows do not correspond to the correct quantitative traits Caution if you are using matrix expansions since this may lead to unwanted configurations as shown below patch_number 3 quanti_nb_trait 5 quanti_selection_model 0 quanti_selection_model_3 1 patch stab sel optima 1 2 3 patch_dir_sel_growth_rate 4 5 6 2 3 7 2 8 3 In this example the growth rate are as follows trait stabilizing selection optima 1 2 3 trait stabilizing selection optima 1 2 3 trait directional selection growth rate 2 8 3 trait directional selection growth rate 4 5 6 trait directional selection growth rate 2 3 7 OU DUO ND HA CHAPTER 5 METAPOPULATION 46 Maybe this behavior was desired but it could also well be that one anticipated the following specification which is wrong trait stabilizing selection optima 1 2 3 trait stabilizing selection optima 1 2 3 trait directional selection growth rate 4 5 6 trait directional selection growth rate 2 3 7 trait
97. rices are defined in whole i e the number of carrying capacities in the 1D matrix example above meets the number of patches 4 patches However if there is a repeating pattern in the matrix it is also possible to define only the repetition In this case the matrix will be repeated as needed For instance the 1D matrix above could also be written in one of the following ways as there is a repetition in it patch_number 4 patch_capacity 20 10 CHAPTER 2 USING QUANTINEMO 10 Note that rows and or columns are repeated as needed If the number of columns or rows is not an entire subset a warning will be returned and the simulation will adjust the matrix as patch_number 5 patch_capacity 20 10 In this example we have 5 patches however the carrying capacities are only specified for 2 patches As 2 is not an entire subset of 5 a warning will be returned and the patches will have the following carrying capacities 20 10 20 10 20 If all values of a matrix are identical one may leave out the brackets In the following example all three declarations result in the same simulation patch_number 4 patch_capacity 20 20 20 20 patch_capacity 20 patch capacity 20 Unbalanced matrices Usually a matrix is balanced i e each row has the same number of columns How ever some parameters such as the parameter quanti_loci_positions allow to have unbalanced matrices i e rows do not necessarily cont
98. rom 1 to quanti_all col_allele2 This keyword specifies the column containing the index of the allele with the larger allelic effect This column is mandatory The index of the allele goes from 1 to quanti_all col dominance This keyword specifies the column containing the dominance effects For any pair of alleles for which the dominance effect is not speci CHAPTER 6 QUANTITATIVE TRAITS 54 fied a dominance effect will be drawn from the normal distribution defined by the parameters quanti_dominance mean and quanti_dominance_var Note that as in all input files for quantiNEMO it is possible to add comments also in the file information box using the hash character or any text quanti_dominance_mean decimal default 0 This parameter allows to specify the mean of the normal distribution from where the dominance effects are randomly drawn Note that a has the smaller allelic effect than az This parameter is only taken into account if the domi nance effects are not defined explicitly by the dominance file quanti_dominance_var decimal default 0 This parameter allows to specify the variance of the normal distribution from where the dominance effects are randomly draw This parameter is taken into account only if the dominance effects are not defined explicitly by the dominance file 6 1 3 Epistatic effects It is possible to simulate epistatic effects between alleles at different loci The followi
99. s and the range is down regulated if the number of alleles is lower Note that in this later case the variance of the allelic effects in a population CHAPTER 6 QUANTITATIVE TRAITS 60 lt Does the allele mutate gt gt No Yes lt Mutation model gt RMM TT IMM A Are the mutation Me the mutation lt _ probabilities defined gt lt _ probabilities defined a a explicilty 27 Sa explicity 7 Yes 7 No Allele ap drawn following Allele are drawn in a Y Allele ane drawn following Allele apy drawn in a the given probabilities normal distribution the given probabilities normal distribution Needed parameters Needed parameters Needed parameters Needed parameters quanti mutation models 0 quanti mutation models 0 quanti mutation model 1 quanti mutation models 1 aguanti mutation rate quanti_mutation_rate quanti_mutation_rate quanti_mutation_rate quanti_allelic_ file guant allelic var guant allelic file quanti_allelic_ var 4 with col_mut_freq 11 PA with col_mut_freq i s we LT Pa in Pa pa a Brow T But Brew T og Bret Figure 6 1 Schematic representation of the mutation process for a specific allele CHAPTER 6 QUANTITATIVE TRAITS 61 may not meet the specified variance Where is the square root of the variance of the normal distribution from where the allelic effects are drawn randomly see parameter quanti_allelic var The probability to mutate to a given allele given that there is a m
100. s replicates file generic_name_var txt 5 None No summary statistics are written The life cycle event Statistics is skipped Whenever the summary statistics are output parameter stat_save not set to 3 a file named generic name legend txt containing a small description of the summary statistics is also generated stat_log_time integer temporal default 1 This is the time interval at which summary statistics are recorded The interval must range between 1 and the number of generations Since the parameter may change over time temporal parameter the summary statistics may be computed for any generation stat_log_time 1 1 10 10 100 100 In this example for the first 9 generations the summary statistics are computed every generation from the tenth until generation 99 they are computed at every tenth generation and from the generation 100 every hundred generation CHAPTER 3 LIFE CYCLE 27 stat dir string default stat This parameter is used to specify a subdirectory within the simulation folder parameter folder where the summary statistic files will be stored If the parameter is not set the files will be stored in the simulation folder string matrix This parameter allows to specify the summary statistics to be computed The arguments are key words standing for one or multiple summary statistics If multiple key words are passed they have to be written as a matrix with
101. simulation such as the elapsed time for the simulation and the replicates and the time of the start and end of a simulation By default quantiNEMO will save all this information in a file named quantinemo log logfile_type 0 1 2 default 0 For each simulation a log file is created containing the settings of the simu lation The file is created for each simulation and is stored in the simulation folder parameter folder The name of the log file is composed of the base name parameter filename and the suffix log There are three possibilities to store this file 0 as input The file contains the same parameters as the settings file plus the parameter seed 1 minimal The file contains a minimal set of parameters still able to per form the same simulation All parameters of the settings file which were set to the default value are not reported 2 maximal The file contains all parameters including the ones not set in the settings file For each parameter the type the default value if it is a temporal parameter and a possible range limit is added as comment seed integer matrix default 1 This parameter specifies the seed which will be used to initialize the random number generator It is possible to pass a single number as seed in the range CHAPTER 4 SIMULATION 36 of O to 4 294 967 295 depending on the computer system corresponding to an unsigned long in C To increase the number of possible seeds
102. ss of quantitative trait t quanti selection model 0 1 2 default 0 This parameter allows to define the selection model for each quantitative trait Section 5 describes how to set the selection pressure at the patch level 0 stabilizing selection Stabilizing selection acts on the quantitative trait The fitness W of quantitative trait t is computed using the following standard Gaussian function for stabilizing selection 6 et Owe W CHAPTER 6 QUANTITATIVE TRAITS 65 Zop is the selection optimum of the current trait and patch parameter patch_stab_sel_optima P is the phenotypic value of the individual and w is the intensity of the selection parameter patch stab sel intensity 1 directional selection Directional selection acts on the quantitative trait A generalized logistic curve Richards 1959 is implemented in quantiNEMO to characterize the directional selection pressure Wi l s e Pragas PY Where r is the growth rate parameter patch dir sel growth rate P is the phenotype with the maximal slope parameter patch_dir_sel_max_ growth and s defines the symmetry of the slope parameter patch dir sel symmetry slope is symmetric by default value 0 5 2 neutral selection The quantitative trait is not under selection The fitness of this quanti tative trait is set to 1 not taking into account the phenotype W 1 6 6 Genetic map quantiNEMO has an underlying genetic map which ma
103. ssible alleles parameter quanti_all Note that for both mutation models the number of alleles have to be odd if the allelic effects and the probabilities to mutate to the alleles given that there is a mutation are set automatically In this case also a warning will be drawn if the number of alleles is below 200 informing that this number of alleles may not well represent the normal distribution of the allelic effects quanti_mutation_rate decimal matrix default 0 This parameter specifies the mutation rate per locus and generation If the argument is a single value the mutation rate for all loci is the same By passing a matrix of mutation rates it is possible to set the mutation rate for each single locus individually By default no mutations occur quanti mutation shape integer default 0 This parameter allows to specify the shape of the gamma distribution from CHAPTER 6 QUANTITATIVE TRAITS 62 which the mutation rates for each locus are drawn The mean mutation rate is given by the parameter quanti_mutation_rate D shape mut_rate shape Where T is the gamma function requiring two parameters The first parameter defines the shape of the gamma distribution parameter guanti mutation shape and the second parameter the scale to the gamma distribution The scale pa rameter is set such does the mean of the gamma distribution is equal to the mean mutation rate mut_rate respectively parameter quanti_mutation_rate If the shape
104. ssible to bypass the end of a line and to write an argument on several lines Note that after the backslash any text on the line is removed 2 3 4 Macros quantiNEMO has two input macros The macros help to keep the input file clear and small In principle the macros just write out the sequence or the repetition before the input file is read by quantiNEMO Sequence With the command seg from to steps it is possible to specify a sequence of data points seq works similar to seq in the statistical package R The macro needs three arguments separated by a comma the first one is the first data point of the sequence the second argument is the last data point of the sequence and the third argument specifies the number of data points including the edges Example patch_capacity seq 100 1000 10 patch_capacity 100 200 300 400 500 600 700 800 900 1000 Both specifications of the carrying capacities are identical i e the macro seq trans lates its arguments to the lower specification The command seg can also be used for temporal parameters with a single number as argument patch_capacity seq 0 100 90 1000 10 patch_capacity 0 100 10 200 20 300 30 400 40 500 50 600 60 700 70 800 80 900 90 1000 CHAPTER 2 USING QUANTINEMO 8 Both specifications of the linear increase of the carrying capacities over time are identical Again the macro seq translates its arguments into the lower specificatio
105. st value of a pair specifies the time i e before which generation the change happens The second value is the new argument A parameter may change as often as any generation A simulation starts at generation 1 Therefore the first change has to have the time value 1 otherwise the parameter cannot be initialized leading quantiNEMO to return an error A temporal argument may be written over several lines patch capacity 1 100 50 200 100 500 Here the carrying capacity is 100 for the first 49 generations 200 from generation 50 on and 500 from generation 100 on 2 3 7 External files In general arguments are written directly after the parameter name on a single line However arguments may be sometimes large e g big matrices temporal parameters leading to poorly readable settings file Using external files for large arguments allows to keep the settings file clear and well readable An external file may be used for any parameter Instead of writing the argument directly after the parameter name the name of an external file is written after the parameter name In order for quantiNEMO to recognize the argument as a file name the prefix has to be added before the file name The external file must contain the argument of the parameter Only a single argument per external file is possible however the same external file may be used for several parameters The format of the argument in the external file follows the sam
106. stant This is the same as model 1 but the environmental variance is readjusted at each generation Thus the narrow sense heritability remains constant over time but not the environmental variance 3 Vg defined by the broad sense heritability Vz constant The vari ance of the environment Vz is set at the beginning of a simulation gen eration 1 and is based on the broad sense heritability H parameter quanti_heritability and the genetic variance Va at generation 1 Note that in this case the environmental variance remains constant over time but not the heritability 4 Vp defined by the broad sense heritability A constant This is the same as model 3 but the environmental variance is readjusted at each generation Thus the broad sense heritability remains constant over time but not the environmental variance quanti_heritability decimal matrix default 1 This parameter depends on the environmental model chosen parameter quanti_environmental_model If Vg is directly set model 0 this parameter is the environmental variance For the environmental model 1 and 2 this pa rameter is the narrow sense heritability h h V4 Vp For the environmental model 3 and 4 this parameter is the broad sense heri tability H H Vg Vp Note that a heritability narrow and broad sense of 0 no genetic component makes no sense for models 1 to 4 in this simulation framework Therefore a value of 0 is not accepted by
107. tch mean coancestry between patch i and j all pairwise com binations computed Genetic diversity n adlt off gendiv n adlt off nbAll n adlt off meanAll n adlt off nbFixLoc n adlt off meanFixLoc n adlt off ho n adlt off hs n adlt off ht A ee N ON N n adlt off nbAll p n adlt off nbFixLoc_p n adlt off ho p n adlt off hs_p mean number of alleles per locus across the entire metapop ulation mean number of alleles per locus and patch number of fixed loci across the entire metapopulation mean number of fixed loci per patches observed heterozygosity following Nei and Chesser 1983 expected heterozygosity following Nei and Chesser 1983 total expected heterozygosity following Nei and Chesser 1983 mean number of alleles per locus within patch computed for each patch number of fixed loci within patch computed for each patch observed heterozygosity within patch i following Nei and Chesser 1983 computed for each patch expected heterozygosity within patch 7 following Nei and Chesser 1983 computed for each patch F statistics following Nei and Chesser 1983 n adlt off fstat n adlt off fst n adlt off fis n adlt off fit global F ST global F ie global Frr Table 7 1 continued on next page CHAPTER 7 NEUTRAL MARKERS 87 Stat name Description n adlt off fst_pair pairwise Fsr between patch i and j all pairwise combina tions c
108. tch stab sel optima 1 patch stab sel intensity 1 In this example the environment consist of two patches with varying selection pres sures Three quantitative traits are simulated The first trait has a selection opti mum at 0 1 in patch 1 and at 0 1 in patch 2 The selection optimum of the second trait is the same in both patches 0 2 The third trait has an optimum at 0 3 in patch 1 and at 0 3 in patch 2 The intensity of the selection is identical for all three traits and in both patches 5 2 2 Directional selection Directional selection may act on quantitative traits The fitness W of a quanti tative trait is computed using the following generalized logistic function Richards 1959 max min W min GET erPryes P Us Where min is the lower asymptote parameter patch dir sel min maz is the upper asymptote parameter patch dir sel max r is the growth rate parameter patch dir sel growth rate Prato is the phenotype with the maximal slope parameter patch dir sel max growth and s defines the symmetry of the curve parameter patch dir sel symmetry the curve is symmetric by default value 1 patch dir sel min patch dir sel min fem CHAPTER 5 METAPOPULATION 43 1 0 0 8 fitness 0 4 0 0 4 2 0 2 4 phenotype patch dir sel min mal decimal matrix temporal default 0 These parameters allow to set the lower asymptote of the selection curve
109. tended Same as point 1 but the file contain the following six additional columns the age class 1 offspring 2 adult the sex 0 male 1 female the ID of the individual the ID of the mother the ID of the father and the fitness of the individual The ID is a unique identifier for each individual of a simulation in the format 345_23 meaning that this is the 345th individual born in patch 23 The IDs of the individual the mother and the father allow to extract pedigree informations if the output is stored for each generation and also to investigate the migration behavior of the individual and its parents CHAPTER 6 QUANTITATIVE TRAITS 71 An example of such a file quanti_save_phenotype is set to 2 2 5 phenotypic_value_trait 1 phenotypic_value_trait 2 phenotypic_value_trait 3 phenotypic_value_trait 4 phenotypic_value_trait 5 1 0 0493 3 203 2 441 0 0683 3 199 2 1 10 1 1 1 O_1 0 345 1 0 4924 3 803 0 869 2 002 2 594 2 1 11_1 8 1 2 2 0 334 1 2 2342 2 931 0 725 0 750 0 698 2 1 12 1 5 2 5 1 0 123 2 0 8623 0 6525 0 857 1 7483 4 194 2 1 16 2 9 2 3 2 0 999 2 1 7752 2 223 3 117 0 3409 2 003 2 1 17 2 3 2 9 2 1 000 2 0 2081 2 803 0 146 0 456 5 137 2 1 11 1 8 1 9_1 0 678 The first line contains the number of patches 2 patches here and the number of traits 5 The next five lines contain the five trait names The following lines contain th
110. ters assuming one selection type after the other that all quantitative traits have the same type of selection assuming for example first that all quantitative traits are under stabilizing selection then in a second step assuming that all quantitative traits are under directional selection This detail is important to understand since CHAPTER 5 METAPOPULATION 45 this makes it clear how a matrix is treated i e how a matrix is expanded if needed Of course finally the selection pressure parameters are only set where needed i e if the type of selection for a given quantitative trait requires the parameter In other words the matrix of a selection pressure has to have the number of rows of the total number of traits and not only of the traits with the corresponding selection pres sure this is controlled by quantiN EMO returning a warning if not met If multiple quantitative traits are simulated with varying selection pressures it is handy to use the row indicator for the rows of the matrix Example patch_number 3 quanti nb trait 5 quanti_selection_model 0 quanti_selection_model_3 1 patch_stab_sel_optima 1 1 2 3 2 2 3 patch dir sel growth rate 3 4 5 6 4 7 8 5 2 3 2 In this example the first two quantitative traits are under stabilizing selection whereas the three last quantitative traits are under directional selection Using the row indicator it is possible to set the selection pressure directly for the
111. the start of a simulation All individuals are fixed for a single allele which is the middle allele i e the allele with the index ntrl_all 2 6 4 Multiple traits quantiNEMO allows to simulate several quantitative traits simultaneously Each trait has its own architecture and may be under different selection pressures quanti nb trait integer default 1 This parameter defines the number of quantitative traits Each trait may have its own specifications but it is also possible to specify parame ters for some quantitative traits together If several quantitative traits are used it is possible to address a certain trait by its number For instance to specify a parameter for the fifth trait one has to append a _5 to the parameter name In contrast if for the fifth trait no parameter with the suffix _5 is passed quantiNEMO checks if the parameter is passed for the fourth trait suffix 4 If this is also not the case quantiNEMO checks if the parameter is passed for the third trait suffix _3 and so forth until a parameter is found Note that a parameter without a suffix is the same as the parameter with the suffix _1 This behavior of quantiNEMO allows to specify parameters for a group of traits An example may make it clearer quanti_nb_trait 12 quanti_loci 5 quanti_loci_7 10 guanti all 1 10 CHAPTER 6 QUANTITATIVE TRAITS 64 quanti_all_4 20 quanti_all_7 10 quanti_all_10 20 quanti mutati
112. this event is skipped After each output a script may be launched to process the output 3 4 Aging This life cycle event simply removes the adults present CHAPTER 3 LIFE CYCLE 28 3 5 Regulation offspring This event performs population regulation before dispersal i e at the offspring stage Due to dispersal some patches may be overcrowded This life cycle event allows to regulate the population sizes down to carrying capacity In fact the reg ulation should only be used if there is no regulation at the breeding stage i e if the parameter breed_model is set to 1 keep number 2 fecundity or 3 fecundity stochastic regulation_model_offspring 0 1 default 0 0 no regulation No population size regulation takes place at the offspring stage i e overcrowding can occur 1 random regulation For each patch quantiNEMO regulates the popu lation size down to its carrying capacity if the population size exceeds carrying capacity Individuals are thereby randomly removed No regu lation takes place if the population size is lower than carrying capacity 3 6 Dispersal This life cycle event allows the exchange of individuals between populations The dispersal rates may vary among patches sexes and time Several dispersal models are available see parameter dispersal_model It is also possible to specify a dispersal matrix which will have precedence over other dispersal parameters After dispersal individuals become adults
113. tion or the trait can be neutral The selection pressure may vary in time and space A quantitative trait is at least specified by the number of loci and number of alleles In this minimal definition the genotypic value of the quantitative trait is purely additive and the traits does not undergo any mutation By default only additive genetic effects are simulated and each new population is initiated by assigning random allelic values within the range 1 quanti all to each locus thus assuring a very large initial variance 49 CHAPTER 6 QUANTITATIVE TRAITS 50 6 1 Architecture quanti_loci integer This parameter specifies the number of loci defining the quantitative trait This parameter is mandatory for the simulation of a quantitative trait quanti_all 1 to 256 default 255 This parameter specifies the maximal number of alleles per locus same number for each locus 6 1 1 Allelic effects Each allele has an allelic effect its contribution to the genotype There are two possibilities to define these effects Either they are defined explicitly for each allele using a separate file see parameter quanti_allelic_file or they can be defined by specifying the variance of the normal distribution of the allelic effects see parameter guanti allelic var If the allelic effects are defined in both ways the explicitly defined effects are used Note that all effects have to be defined in the same way i e explicitly or by their distribut
114. tion txt simulation_mean txt simulation_legend txt statistic files statistics for each replicate statistics across replicates statistic legends RH 2 7 Batch mode quantiN EMO allows to perform multiple simulations by executing a single command There are two ways to perform such batch simulations Note that these two types may not be mixed i e be used at the same time 2 7 1 Multiple settings files A normal simulation can be launched by passing the settings file name as parameter to the executable It is also possible to pass several settings file names to the executable In this case a simulation for each settings file is executed consecutively gt quantinemo exe siml ini sim2 ini In this example quantiNEMO is launched with two settings files sim1 ini and sim2 ini The simulations will be executed one after the other leading to the following console output Reading settings file siml ini done Reading settings file sim2 ini done SIMULATION 1 2 SIMULATION 1 2 done CPU time 00 00 15s CHAPTER 2 USING QUANTINEMO 18 SIMULATION 2 2 SIMULATION 2 2 done CPU time 00 00 16s 2 7 2 Sequential parameters A batch simulation may also be launched by a single settings file if so called sequen tial parameters are used Sequential parameters are any parameter with not only one but several arguments Note that a seque
115. traits or neutral markers are specified The simulation generates no output apart from the log file Although this simulation works it makes no sense since no output is generated This can be improved by specifying quantiNEMO to compute summary statistics on the demog raphy generations 500 patch capacity 1000 stat adlt demo This settings file results in exactly the same simulation as the previous one but now three additional files are generated The first file lists the names of the computed summary statistics simulation legend txt while the two other files are almost identical containing the summary statistics for each generation One of these latter files contains the summary statistics for each generation and replicate separately simulation txt while the other one shows the summary statistics averaged across replicates simulation_mean txt CHAPTER 2 USING QUANTINEMO 14 To simulate genetic data one has to add either a quantitative trait or a neutral marker to the simulation This is done by specifying the number of loci to simulate for quantitative traits and neutral markers separately To force quantiNEMO to compute some summary statistics on quantitative traits respectively on neutral markers one has also to specify corresponding summary statistics For example generations 500 patch_capacity 1000 stat adlt demo quanti n adlt fstat quanti_loci 1 ntrl_loci 1 In th
116. trl_all 10 ntrl_mutation_model 0 ntrl_mutation_rate le 4 ntrl_save_genotype 2 ntrl_genot_logtime 10 ntrl_genot_dir ntrl_genotype statistics stat n adlt fstat stat_save 1 stat_log_time 10 stat_dir stats A simulation based on this settings file named quantiNemo_example ini produces the following output to your terminal window N K K K K K K OK K K A OK OK OK K K OK OK K K OK OK OK OK K OK OK K K K K K K OK N K K K OK OK OK K K K OK OK OK K OK K K quantiNEMO EE de EE dd dd dd EE EE EE Ed de III CK Release 1 0 0 0 Jan 01 2008 00 00 00 Copyright C 2008 Samuel Neuenschwander http www unil ch popgen softwares quantinemo OO OK Reading settings file quantiNemo_example ini done SETTINGS Simulation 100 generations 1 replicates Loaded traits Neutral marker type 5 loci 10 alleles Quantitative trait 5 loci 255 alleles stabilizing selection Life cycle sequence 1 breed 2 save_stats 3 save_files CHAPTER 2 USING QUANTINEMO 16 4 aging 5 disperse 6 regulation Metapopulation 10 populations Migration model island Mating system random mating hermaphrodite SIMULATION replicate 1 1 00 00 14 100 100 SIMULATION done CPU time 00 00 15s quantiNEMO terminated successfully This output informs you that this simulation was parameterized by the settings file quantiNemo_example ini The simulation consisted of one q
117. trl_genot_dir 84 ntrl genot logtime 84 ntrl genot script 84 ntrl_genot_sex 85 ntrl_ini_allele model 81 ntrl_ini genotypes 80 ntrl loci 76 ntrl loci positions 82 ntrl_loci_positions_random 83 ntrl_mutation_model 79 ntrl_mutation_rate 79 ntrl_mutation_shape 80 ntrl_nb_trait 81 ntrl_save_genotype 83 overwrite 34 patch_capacity 38 patch_capacity fem 38 patch capacity mal 38 patch_dir_sel_growth_rate 43 patch_dir_sel_growth_rate_fem 43 patch_dir_sel_growth_rate_mal 43 patch_dir_sel_growth_rate_var 44 patch_dir_sel max 43 patch dir sel max fem 43 patch dir sel max growth 43 patch dir sel max growth fem 43 patch dir sel max growth mal 43 patch dir sel max growth var 44 patch dir sel max mal 43 patch dir sel min 42 patch dir sel min fem 42 patch dir sel min mal 43 patch dir sel symmetry 43 patch dir sel symmetry fem 44 patch dir sel symmetry mal 44 90 INDEX patch_dir_sel_symmetry_var 44 patch_ini_size 39 patch ini size fem 39 patch_ini_size_mal 39 patch_number 38 patch_stab_sel_intensity 41 patch_stab_sel intensity fem 41 patch stab sel intensity_mal 41 patch stab sel intensity var 42 patch stab sel optima 41 patch stab sel optima fem 41 patch stab sel optima mal 41 patch stab sel optima var 42 postexec script 36 guanti all 50 quanti_allelic file 50 quanti_allelic_var 52 quanti_dominance file 53 quanti_dominance_mean 54 quanti dominance_var
118. uantitative trait and one type of neutral markers The simulated life cycle is indicated and shows that summary statistics and genotypes or phenotypes are output Only one replicate of 100 generations was performed For each replicate the elapsed time hh mm ss is printed out to the console and at the end of the simulations also the total elapsed time The output of this simulation was stored in the following structure relative to the executable simulation 2008 01 01_00 00 00 automatically named folder simulation log log file quanti_phenotype phenotype folder simulation_g010 phe phenotypes of generation 10 simulation_g020 phe simulation_g030 phe simulation_g040 phe simulation_g050 phe simulation_g060 phe simulation_g070 phe simulation_g080 phe simulation_g090 phe simulation_g1l00 phe quanti_genotype genotype folder quantitative simulation_g010 dat genotypes at generation 10 simulation g020 dat simulation_g030 dat simulation_g040 dat simulation_g050 dat simulation_g060 dat simulation_g070 dat simulation_g080 dat simulation_g090 dat simulation_g1l00 dat CHAPTER 2 USING QUANTINEMO 17 ntrl_genotype genotype folder neutral simulation_g010 dat genotype at generation 10 simulation_g020 dat simulation _g030 dat simulation g040 dat simulation g050 dat simulation_g060 dat simulation_g070 dat simulation_g080 dat simulation_g090 dat simulation g100 dat stats simula
119. ue of the parameter dispersal k min Above the threshold the factor is set to the value of the parameter dispersal_k_max dispersal_k symmetry decimal temporal default 1 This parameter specifies the symmetry of the curve The default value of 1 results in a symmetric curve 3 7 Regulation adults This event performs population regulation after migration Due to dispersal some patches may be overcrowded This life cycle event allows to regulate the population sizes down to carrying capacity In fact the regulation should only be used if there is no regulation at the breeding stage i e if the parameter breed_model is set to 1 keep number 2 fecundity or 3 fecundity stochastic regulation_model_adults 0 1 default 0 0 no regulation No population size regulation takes place at the adults stage i e overcrowding can occur 1 random regulation For each patch quantiNEMO regulates the popu lation size down to its carrying capacity if the population size exceeds carrying capacity Individuals are thereby randomly removed No regu lation takes place if the population size is lower than carrying capacity 3 8 Extinction This event allows to randomly wipe out populations The probability that a pop ulation goes extinct is specified by the parameter extinction_rate If a patch goes extinct all individuals of the patch are removed If the extinction rate is zero this event is skipped extinction_rate decimal default 0
120. us The following equation is used to compute the genotypic value at a given locus Gi a ay kii a _ a l Where G is the genotypic value of locus i a and a are the effects of the two alleles at locus and k is the dominance value between allele and k can have the following effects e k lt 1 underdominance e k 1 allele a is dominant e k 0 no dominance purely additive e k 1 allele a is dominant e k gt 1 overdominance There are two possibilities to define these dominance effects Fither they are defined explicitly for each pair of alleles using a separate file see parameter guanti dominance file or by defining the normal distribution from where the dominance effect are ran domly drawn see parameters quanti_dominance_mean and quanti_dominance_var If the dominance effects are defined in both ways the explicitly defined effects are used Note that all effects have to be defined in the same way i e explicitly or by their distribution By default the parameters guanti dominance mean and quanti_dominance_var are set to zero resulting in a purely additive genotypic value of a locus CHAPTER 6 QUANTITATIVE TRAITS 93 quanti_dominance file string default gt This parameter allows to pass the name of a file containing dominance effects The number of alleles and loci has to be in line with the parameters quanti_loci and quanti_all The information can be set globally for all loci
121. utation is identical to the frequency of this allele defined by the distribution of the allelic effects see parameter quanti_allelic_var i e follow a normal distribution 1 IMM Increment Mutation Model At a given mutation an allele effect is drawn randomly and is added to the current allelic effect Anew Qold Amut Where anew is the effect of the new allele amut is the effect of the drawn allele and asa is the effect of the old allele before the mutation The probability to mutate to a certain allelic effect depends on its probability to mutate to it given that there is a mutation These probabilities and also the effects of the alleles can explicitly be set by the allelic file see section 6 1 1 If this is not the case quantiNEMO allocates the allelic effects and the probabilities to mutate to the alleles given that there is a mutation automatically The effects of the alleles are regularly spaced between 200 and 200 where is the square root of the variance of the normal distribution from where the allelic effects are drawn randomly see parameter quanti_allelic_var The probability to mutate to a given allelic effect given that there is a mutation is identical to its allele frequency i e follow a normal distribution This mutation model requires that the allelic effects are regularly spaced around zero that the number of alleles is odd thus the allele with index quanti_all 2 is zero and that there are at least 51 po
122. y the selection pressure individually for quantitative traits and patches matrices may be used They are adjusted to the number of quantitative traits and to the number of patches if needed see section 2 3 5 4 If a parameter does not change among quantitative traits and patches a single value may be used as argument Selection pressures have to be specified either for each sex separately parameters with the suffix fem for females and _mal for males or for both sexes together param eters without a suffix In the first case both sex specific parameters have to be set if two sexes are simulated In the latter case the selection pressure of females and males are assumed to be identical Each row of the matrix corresponds to a quantitative trait each column to a patch patch_1 patch_2 patch n trait 1 patch_1 patch 2 patch n trait 2 patch_l1 patch 2 patch n trait m quantiNEMO supports two types of selection stabilizing and directional selection The type of selection may vary among quantitative traits but not among patches and sexes Selection pressure may change through time allowing simulating a dynamic environment such as global warming CHAPTER 5 METAPOPULATION 41 5 2 1 Stabilizing selection Stabilizing selection may act on the phenotype P of quantitative traits The selection pressure is defined by the parameters patch stab sel optima Zom and patch stab sel intensit
123. y w The fitness W of a quantitative trait is computed using the following standard Gaussian function for stabilizing selection ot Wee 2 Y CO 2 o o 2 N ao D Y 3 N l o Zopt o T T T T 4 2 0 2 4 phenotype patch_stab_sel_optima patch stab sel optima fem patch stab sel optima mal decimal matrix temporal default 0 These parameters allow to set the selection optimum 20 for each patch and quantitative trait patch_stab_sel_intensity patch_stab_sel_intensity_fem patch_stab_sel_intensity_mal decimal matrix temporal default 1 These parameters allow to set the selection intensity w for each patch and quantitative trait A small value results in a strong selection pressure whereas a large value results in a weak selection pressure CHAPTER 5 METAPOPULATION 42 patch stab sel optima var decimal matrix default 0 This parameter specifies the variance of the normal distribution by which the selection optimum varies between generations e g annual fluctuations of the mean temperature By default the local selection optimum does not vary patch stab sel intensity var decimal matrix default 0 This parameter specifies the variance of the normal distribution by which the selection intensity varies at each generation e g annual fluctuations of the mean temperature By default the local selection intensity does not vary Example patch number 2 guanti nb trait 3 pa
124. y consist of several chromo somes This allows to explicitly position all types of loci on the map quantitative trait loci QTL and neutral markers The unit of the genetic map is centi Mor gans Haldane 1919 cM If the genetic map of a trait is not specified by either of the following parameters the loci for that trait are assumed to be unlinked and independent quanti loci positions matrix default For each trait this parameter allows to specify explicitly the positions of the loci in centi Morgans from the beginning of the chromosome The brackets separate the chromosomes Within a chromosome the positions of the different loci are separated by a space Note that chromosomes may contain different numbers of loci but that the total number of loci must correspond to the parameter quanti_loci It is possible to skip chromosomes if they don t contain loci see section 2 3 5 4 quanti loci 11 quanti_loci_positions 1 10 13 10 40 60 80 100 10 40 60 80 100 CHAPTER 6 QUANTITATIVE TRAITS 66 In the example above the quantitative trait is defined by 11 loci located on three of at least four chromosomes The first chromosome contains a single locus at position 10 cM No loci are located on the second chromosome The third and fourth chromosomes have 5 loci at positions 10 40 60 80 and 100 cM quanti_loci_positions_random matrix default This parameter allows to specify the number of c

quantiNEMO

Contents

Download Pdf Manuals

Related Search

Related Contents