CopyCat – The Cophylogenetic Analysis Tool User Manual
Contents
1. In case the user has selected the above option CopyCat chooses AxParafit_win exe and AxPcoords win exe operating system Windows If you wish to use externally compiled executables e g using the ACML or MKL libraries they should obey to the naming convention lt PROGRAM gt lt OS gt _ lt LIBRARY gt as e g AxParafit_ win MKL exe AxPcoords win MKL exe and moved into the code subdirectory CopyCat will detect and use them for the next computation Such an optimized executable will always be preferred to the default one Once the user has selected the AxPcoords AxParafit option the option correction method used in DistPCoA the second item on the second tab will be disabled because correction methods for negative eigenvalues are not supported by AxPcoords Setup Show setup menu at next program start If you want to change the working directory you just have to check this option On the next start of CopyCat the configuration dialogue will appear again Tutorial A step by step example run of Copycat This tutorial focusses on the kind of input needed for certain steps in Copycat and shows the output produced by that input The underlying data set for this example run is the list of European smut fungi and their hosts from Vanky Van94 Van05 This data set does not contain parasite or host trees or distance matrices so we have to construct them using the NCBI taxonomy 1 The input2. Swofford 1991 as well as information content Thorley and Page 2000 are measures of topological resolution Resolution is just the number of internal nodes divided by the maximum possible number of internal nodes i e n 2 and thus bound between 0 and 1 A value of 1 indicates full resolution The cladistic information content has some theoretical advantages but in case of not fully resolved topologies this measure may rapidly converge towards 0 if the number of taxa increases and thus may not be applicable when dealing with large datasets N B Due to the presumed trade off between the number of taxa and the topological resolution available as input for ParaFit the user has to base her decision whether to conduct a ParaFit analysis based on taxonomic data not only on the number of resolved associations but also on the amount of topological resolution Even though one of the advantages of ParaFit is that it does not require fully resolved binary trees Legendre et al 2002 trees may well display not enough topological resolution too many polytomies to be of value in conducting cophylogenetic analyses As an extreme example consider a totally unresolved taxonomic tree Since in that case the eigenvectors of all taxa as output by DistPCoA will be identical such a tree will lead to all associations being insignificant just for trivial reasons Step 4 create the parasite distance matrix This is the same as above Step 5 vali
3. association file Showed results step 2 statistics select association file compute distribution of parasite host taxa across the systematic divisions show BSD show broken stick distribution exclude overrepresented taxa step 3 filter parasite hostassociations this step also creates the parasites and hosts taxa lists exclude certain systematic divisions NCBI T Bacteria F Invertebrates l Mammals l Phages l Plants l Primates l Rodents M Synthetic M Unassigned F Viruses F Vertebrates M Environmental samples additional filter options leave associations in their current state C compress all associations to the following rank genus vy remove associations containing taxa taxon parasite host taxa to remove apply settings to association list elcome to CopyCat The cophylogenetic analysis tool INFO The operating system is WINDOWS 2000 INFO The working directory is E Eclipse workspace Copycat default wdin ID_8 INFO In order to change the configuration check show setup menu at next program start in the Setup menu Y Figure 2 The first tab of CopyCat CopyCat The first tab The first tab see figure 2 deals with the creation and pre processing of an association table containing parasite or host associations Finding NCBI taxonomy IDs Given an unresolved association list containing parasite or host names the resolve association list item
4. by the purple WRAPPER tag The wrapper is finished once the message window contains the following three information lines at the very end here example values INFO on host tree Resolution 0 252525 INFO on host tree Balance input is no rooted binary tree INFO on host tree Information content Infinity The working directory should now contain the files hosts dist and host out tree We repeat the procedure for the parasites filtered using option 0 txt file by selecting the distance matrix from host taxa list option in the box below The distance matrix is finished once the following three lines appear at the end of the message window here example values INFO on parasite tree Resolution 0 252525 INFO on parasite tree Balance input is no rooted binary tree INFO on parasite tree Information content Infinity The working directory should now contain the files parasites dist and parasites out tree These info messages show the resolution and the phylogenetic information content which can t be computed for large numbers of taxa of the host parasite tree created If the resolution is not satisfying enough you have the option to cancel the further cophylogenetic analysis at this stage If you want to proceed you have to choose the validate button The validation step ensures that the taxa contained in the following three files are consistent with respect to each other This means that a parasite s n
5. in that specific Parafit run These files should reside in your working directory Once the three files Hostpara out parasite distance matrix and host distance matrix have been specified Copycat opens a new window showing the results An example is shown in illustration 2 18 Parafit results significant links are coloured gray else remain white Entyloma_dactylidis_63287 358 Festuca_rupicola_208425 0 01400 0 01400 359 Entyloma_dactylidis_63287 Poa_annua_93036 0 01200 0 01200 360 Entyloma_dactylidis_63287 Poa_palustris_269377 0 01100 0 01100 361 Entyloma_dactylidis_63287 Poa_compressa_269373 0 01600 0 01600 362 Entyloma _dactylidis 63287 Poa_alpina_87701 0 01200 0 01200 363 Entyloma_dactylidis_63287 Poa_bulbosa_202404 0 01400 0 01400 364 Entyloma_dactylidis_63287 Poa_4544_4544 0 01800 0 01800 365 Entyloma_dactylidis_63287 Lolium_perenne_4522 0 01400 0 01400 366 Entyloma_dactylidis_63287 Cynosurus_cristatus_211939 0 01000 0 01000 367 Entyloma_dactylidis_63287 Cynosurus_echinatus_211940 0 01000 0 01000 368 Entyloma_dactylidis_63287 Lamarckia_aurea_29682 0 01200 0 01200 369 Entyloma_dactylidis_63287 Puccinellia_distans_13649 0 01200 0 01200 370 Entyloma_dactylidis_63287 Deschampsia_cespitosa_37723 0 01500 0 01500 371 Entyloma_dactylidis_63287 Dactylis_glomerata_4509 0 01400 0 01400 372 Entyloma_dactylidis_63287 Trisetum_flavescens_87477 0 02000 0 02000 ia 373 Entyloma_dactylidis_63287 Trisetum_spicatum_100813 0 02800 0 02800 374 Ent
6. of the association table The user hits the select association file button and then specifies an association table AT refer to first step The AT is read and the following features are reported to the message window The distribution of the taxonomical ranks within the set of hosts and the set of parasites respectively in accordance to the NCBI taxonomy The affiliation of each parasite and host taxon respectively to one of the following divisions division 0 Archaea division 1 Bacteria division 2 Eukaryota division 3 Viroids division 4 Viruses division 5 Other division 6 Unclassified division 7 taxon ID not found in NCBI taxonomy The number of associations contained in the AT The number of different parasite taxa in AT The number of different host taxa in AT The estimated size of the association matrix drawn from the AT N B As a rule a phylogenetic tree derived from character data e g molecular sequences contains more information than a taxonomy of the same taxa since the topology of the latter is usually much less resolved However if a cophylogenetic study is based on specific marker sequences such as 16S rRNA or ITS it is limited to species for which there is a common marker gene available Even though the number of single locus or even genome sequences is steadily increasing we presume that NCBI taxonomic data are available for many more taxa than are such orthologous genetic data We guess this rule also
7. patristic distance matrix results in which each pairwise distance between two taxa A and B represents the sum of the number of taxa in which A is contained including A but not B and the number of taxa in which B is contained including B but not A Accordingly by default branch lengths in the NCBI taxonomy tree may be larger than one representing more information extracted from the taxonomy than just the topology Use AxPcoords and AxParafit instead of DistPCoA and Parafit By selecting this option AxPcoords and AxParafit are used instead of Legendre s programs Legendre et al 2002 Originally CopyCat was designed for supporting the programs DistPCoA computation of principal coordinates from a distance matrix and Parafit providing tests for both overall phylogenetic congruence as well as for the significance of individual host parasite associations As of April 2007 CopyCat supports Alexandros Stamatakis programs AxParafit and AxPcoords which are highly optimized versions of Parafit and DistPCoA respectively CopyCat searches within the code subdirectory which contains all executables for the analyses By default this folder contains the following 13 operating system dependant executables parafit_win exe distpcoa_win exe AxParafit_win exe AxPcoords win exe The attribute win denotes the executables for the Windows platform accordingly mac for Macintosh and linux for Linux systems are used
8. 0 Entyloma_dactylidis_63287 Poa_palustris_269377 0 01100 0 01100 361 Entyloma_dactylidis_63287 Poa_compressa_269373 0 01600 0 01600 362 Entyloma _dactylidis 63287 Poa_alpina_87701 0 01200 0 01200 363 Entyloma_dactylidis_63287 Poa_bulbosa_202404 0 01400 0 01400 364 Entyloma_dactylidis_63287 Poa_4544_4544 0 01800 0 01800 365 Entyloma_dactylidis_63287 Lolium_perenne_4522 0 01400 0 01400 366 Entyloma_dactylidis_63287 Cynosurus_cristatus_211939 0 01000 0 01000 367 Entyloma_dactylidis_63287 Cynosurus_echinatus_211940 0 01000 0 01000 368 Entyloma_dactylidis_63287 Lamarckia_aurea_29682 0 01200 0 01200 369 Entyloma_dactylidis_63287 Puccinellia_distans_13649 0 01200 0 01200 370 Entyloma_dactylidis_63287 Deschampsia_cespitosa_37723 0 01500 0 01500 371 Entyloma_dactylidis_63287 Dactylis_glomerata_4509 0 01400 0 01400 372 Entyloma_dactylidis_63287 Trisetum_flavescens_87477 0 02000 0 02000 E 373 Entyloma_dactylidis_63287 Trisetum_spicatum_100813 0 02800 0 02800 374 Entyloma_dactylidis_63287 Agrostis_canina_218142 0 02300 0 02300 375 Entyloma_dactylidis_63287 Agrostis_capillaris_204232 0 02400 0 02400 376 Entyloma_dactylidis_63287 Agrostis_stolonifera_63632 0 02100 0 02100 377 Entyloma_dactylidis_63287 Agrostis_alba_29659 0 02600 0 02600 378 Entyloma_dactylidis_63287 Phleum_pratense_15957 0 02100 0 02100 379 Entyloma_dactylidis_63287 Alopecurus_pratensis_15304 0 02900 0 02900 380 Entyloma_dactylidis_63287 Holcus_lanatus_29679 0 02300 0 02300 381 Entyloma_dactylidis_6328
9. 06 07 07 CopyCat The Cophylogenetic Analysis Tool User Manual CopyCat is a software tool written in Java which provides an easy and fast access to cophylogenetic analyses It incorporates a wrapper for the program ParaFit which conducts statistical tests for the presence of global congruence between host and parasite phylogenies and for the significance of individual host parasite associations CopyCat offers various features such as the creation of customized host parasite association data the reconstruction of host or parasite trees from the NCBI taxonomy and the computation of several tree statistics As of April 2007 CopyCat supports Alexandros Stamatakis programs AxParafit and AxPcoords which are highly optimized versions of Parafit and DistPCoA respectively see section CopyCat Available menu bar options This manual describes the features specific to CopyCat regarding the principles of the statistical tests implemented in ParaFit users are strongly advised to consult Legendre et al 2002 The literature cited in the references section is also suggested for further reading If you use CopyCat you should cite the accompanying paper Meier Kolthoff J P Auch A F Huson D H G ker M COPYCAT Co phylogenetic Analysis tool Bioinformatics 2007 Epub ahead of print Check list After installing CopyCat please make sure that you have verified the following items You have either a Win32 compatible or a Linux bas
10. 7 Avenula_pubescens_87471 0 02300 0 02300 382 Entyloma_ossifragi_272710 Narthecium_ossifragum_114204 0 70600 0 70600 383 Rhamphospora_nymphaeae_62645 Nymphaea_alba_34301 0 30900 0 30900 384 Doassinga_callitrichis_62640 Callitriche_stagnalis_119595 0 01500 0 01500 x the pre defined significance value 0 02 apply Overall cophylogenetic structure is highly significant 0 00100 lt 0 02 sig val 538 links out of 634 are significant dump results to working directory Figure 7 CopyCat s representation of the ParaFit output By selecting dump information to working directory the results are stored in a simple text file ASCII format Lines contained in that file marked by a character represent significant links while lines starting with a character are considered as non significant CopyCat Available menu bar options The menu bar offers the following options File Transfer content of working directory to a place of your choice Once your work with CopyCat is done you might want to transfer data from the working directory to a directory of your choice This can be done by selecting this option Download NCBI taxonomy file s As mentioned above it is advisable to get the latest NCBI taxonomy file from time to time The taxonomy itself is steadily improved due to the incorporation of more recent phylogenetic insights and the total number of both terminal taxa and taxa of higher rank included in the taxono
11. An association table of smut fungi parasites and their respective hosts Here is an excerpt of the input file smut fungi association table txt 3 AMD Core Math Library 4 Intel Math Kernel Library 14 Anthracoidea altera Carex fuliginosa Anthracoidea angulata Carex hirta Anthracoidea arenaria Carex arenaria Anthracoidea arenaria Carex brizoides Anthracoidea arenaria Carex ligerica Anthracoidea arenaria Carex ovalis Anthracoidea arenaria Carex praecox Anthracoidea aspera Carex appropinquata Anthracoidea aspera Carex chordorrhiza Anthracoidea aspera Carex diandra Anthracoidea aspera Carex glareosa Anthracoidea baldensis Carex baldensis Text 1 The unresolved association table Parasites are in the left column hosts in the right column Each parasite is separated from its respective host by a tab character The file contains 1853 host parasite associations 2 Resolving the association table Each parasite host contained in the NCBI Taxonomy Database has a unique taxonomy ID This step tries to gather these IDs Each association containing both a parasite and a host with a valid ID is used in the resulting so called resolved association table A representation of this resolving step is shown below see illustration 1 15 Resolved association list 645 1853 associations resolved light grey association resa no parasites O 4 Anthracoidea_arenaria_265863 Carex_arenaria_234466 5 A
12. ame contained in the association table should exist in the parasite distance matrix and so forth If the validation is successful you will have the option to immediately start the Parafit run on this machine or as an alternative to pack all relevant files to a zip archive This archive can be transferred to another probably more powerful machine Once the ParafitWrapper has started either on this or another machine it generates a random session ID similar to Copycat and creates a respective subdirectory within your working directory This directory is named after that session ID The time the wrapper is running any interaction with Copycat is blocked Depending on the size of the input data the Parafit run can be a time consuming issue 5 Analysis of the Parafit results Parafit performs tests for both the overall phylogenetic congruence as well as the significance of individual associations 17 These results are listed in the file HostPara out Copycat reports the location of HostPara out by printing WRAPPER Please check file D Copycat defaultwdir ID_3738 ID_8313 Hostpara out for results HostPara out holds the results on the individual links in the following format Parasite 17 Host154 F1 Prob1 0 94600 F2 0 00016 Prob2 0 99200 As Parasite 17 and Host 154 provide no information on the actual organisms Copycat needs to know the location of the parasite and the host distance file used
13. at deFaultwair TAXDMPLOCATION D copycat input data taxdmp zip IV Do not enter this setup menu at the next program start apply save and proceed exit copycat Figure 1 CopyCat s configuration dialog Here you might want to choose another working directory WORKING DIR or another directory for your custom data USERDATA The latter will be relevant if CopyCat asks for a file to open then this directory will be prompted first To proceed please click apply followed by save and proceed By pressing exit copycat the whole application will close If the configuration is not complete the dialogue will be displayed again until everything is well configured Now the CopyCat application launches It generates a random session ID here ID 9511 and creates a subdirectory for this session within your working directory e g D CopyCat myworkingdir ID 9511 All files and results are now going to be stored in that directory The contents of that folder and thus the results of this session can be examined by selecting the View menu item 2The following screen shots were made using the Windows version of CopyCat the look amp feel of the Linux version differs slightly CopyCat The cophylogenetic analysis tool this session has ID 8 Eq File View Options Setup Help Configuration and Execution of Parafit Evaluation of Parafit Results r step 1 resolve association file load
14. ciation matrix to be conducted by ParaFit can be specified ParaFit requires principal coordinates inferred from host and parasite distance matrices as computed by e g the program DistPCoA Legendre and Anderson 1998b DistPCoA supports two correction methods for negative eigenvalues Lingoes method and Cailliez method The user might choose one of these or simply select the no correction method option The advantages and disadvantages of the several corrections methods are explained in e g Legendre and Anderson 1998a and Legendre and Legendre 1998 pp 432ff Step 2 The user selects an association file Step 3 create host distance matrix The user has three possibilities The user selects a pre existing host distance matrix The user specifies a list of host taxa which is used to reconstruct the NCBI host tree The tree is automatically converted to a patristic distance matrix Farris 1967 The user specifies a host tree which is used to create the distance matrix Once a tree is constructed several of its features are reported to the message window Balance is a measure of tree balance as described by Colless 1982 Note that balance can only be computed for rooted binary trees Cherry is the measure of tree balance suggested by McKenzie and Steel 2000 divided by the maximum possible number of cherries n 2 provided that n is the number of taxa Resolution described as Colless consensus fork index in
15. dation If all input files have been specified and all parameters been set the user should then hit the validate the specified data button During this validation 9 the program reports whether the taxon names in the association matrix are consistent with the taxon names in the host and parasite distance matrix or not If taxon names are present in the association table which are not contained in the respective distance matrix the program returns an error In the opposite case taxon names from the distance matrix can t be found in the association table the program offers a shrink distance matrix option which allows the removal of the respective columns and rows from the distance matrix N B In case host or parasite trees are derived by pruning from larger phylogenies it is much more convenient to change just the association table than to manipulate the trees themselves This feature of CopyCat results in a great gain in user flexibility with respect to running ParaFit with slightly different sets of taxa If the validation returns no errors the following two options are enabled start local analysis The ParaFit wrapper is started with the parameters specified above prepare data for remote analysis As an alternative all input files the ParaFit wrapper including ParaFit and DistPCoA and a setup file are put into a ZIP archive The archive can be transferred to a high performance machine After archive extraction t
16. e 3 An association table as resolved by CopyCat The entries coloured in light grey have for both parasite and host a valid ID assigned as provided by the NCBI Taxonomy Database If for both parasite and host name a respective taxon ID is retrieved this parasite host association appears in the final resolved association table The set of all host or parasite taxa contained in that table is used for the reconstruction of the NCBI host or parasite tree respectively If a certain parasite or host name can t be resolved e g because of a misspelling the user has the possibility to manually enter a proper NCBI taxonomy ID The format for this is the organism name followed by the taxonomy ID e g Homo sapiens 9606 The user should ascertain that this taxonomy ID is definitely contained in the NCBI taxonomy database The better and most consistent way of retrieving a correct ID is as follows The user corrects the organism name and then applies the changes by pressing the apply button By entering a name available in the NCBI taxonomy the respective organism field in the association table should then yield the corresponding ID Finally all resolvable associations are written to a file after having selected the dump results to working directory button option Along the new association file two files containing only the parasites and respectively only 4 the hosts are written to the same directory Statistics Characteristics
17. ead parasites The broken stick method provides an objective means to distinguish these from normal species Filtering of an association table This step allows the selection of divisions as defined within the NCBI taxonomy whose corresponding taxa and hence associations should be completely removed from the table The additional filter option provides a so called rank mapping feature Each taxon is mapped to its respective parent taxon until the specified rank e g genus or family is obtained Redundant entries are removed automatically The field remove parasite host associations containing specific parasite host taxon IDs allows the removal of associations whose taxa are listed in this box Valid input is a list of space separated taxa IDs each of them with the leading letter P or H indicating the taxon s membership in the group of parasite or host taxa the taxa being selected via the BSD option are listed in this field For example H9606 would result in a removal of those associations having a host associated with the taxonomy ID 9606 Homo sapiens 9606 would also be a valid input although only the ID is of interest CopyCat The second tab This tab see figure 5 deals with the selection of the input files and parameters for ParaFit Legendre et al 2002 and with the ParaFit run itself Here the most important feature is a wrapper for the program ParaFit preparing and finally providing the properl
18. ed operating system currently tested under Windows 2000 Windows XP and Linux x86 32bit with GTK 2 0 Your machine is equipped with at least 512 MB of memory Even though it is not recommended hardware with less memory can sometimes be used as long as the NCBI taxonomy facilities are not applied The Java 1 5 runtime environment is installed by entering the command java version in a command console the currently installed version is reported and the Java Binary must be included in the PATH environment variable this is done automatically by the Java Installer for Windows Java 1 5 a k a JDK5 is available here http java sun com javase downloads index jsp You have downloaded the current NCBI taxonomy file taxdmp zip from the following URL ftp ftp ncbi nih gov pub taxonomy taxdmp zip and placed this file in the input data subfolder It is located in your CopyCat installation folder e g home john CopyCat input data As the NCBI taxonomy is updated on a regular basis it is advisable to get the latest version of taxdmp zip from time to time 1 Of course any kind of associations with hosts can be examined in that way including mutualists We refer to parasites only just for convenience The first start Once you have started CopyCat for the first time the following configuration dialogue will appear copycat configuration menu USERDATA Ix D copycat userdata WORKING_DIR D copyc
19. graphs 69 1 24 Legendre P and Anderson M J 1998b Program DistPCoA Departement de sciences biologiques Universite de Montreal 10 pages Available at http www bio umontreal ca Casgrain en labo distpcoa html Legendre P and Legendre L 1998 Numerical Ecology Second English Edition Elsevier Amsterdam 853 XV pages Legendre P Desdevises Y and Bazin E 2002 A statistical test for host parasite coevolution Systematic Biology 51 2 217 234 McKenzie A and Steel M 2000 Distributions of cherries for two models of trees Math Biosci 164 81 92 Swofford D 1991 When are phylogeny estimates from molecular and morphological data incongruent Pp 295 333 in Miyamoto M M and Cracraft J Eds Phylogenetic analysis of DNA sequences Oxford University Press New York Oxford 19 Thorley J L and Page R D 2000 RadCon phylogenetic tree comparison and consensus Bioinformatics 16 5 486 487 Vanky K 1994 European smut fungi Gustav Fischer Verlag Stuttgart Jena New York Vanky K 2005 European smut fungi Ustilaginomycetes p p and Microbotryales according to recent nomenclature Mycologia Balcanica 2 169 177 2005 20
20. he wrapper can be invoked by the command java Xmx512M jar ParafitWrapper jar The Xmx switch denotes the maximum amount of memory the wrapper has for its own disposal here 512 MB CopyCat The third tab This tab see figure 6 deals with the evaluation of the ParaFit results 10 r CopyCat The cophylogenetic analysis tool this session has ID 8 _ x elcome to CopyCat The cophylogenetic analysis tooll NFO The operating system is WINDOWS 2000 INFO The working directory is E Eclipse workspace Copycat default wdinID_8y INFO In order to change the configuration check show setup menu at next program start in the Setup menu Figure 6 The third tab of CopyCat After the analysis has ended an output file called Parafit out should have been created In this step this output file is specified via the open dialogue together with the host and parasite distance matrix used in that ParaFit run The distance matrix files are needed to display the correct organism names instead of the non interpretable labels like Parasite 4 or Host 17 A sample ParaFit output as resolved by CopyCat is shown in figure 7 11 358 Entyloma _dactylidis_ 63287 Parafit results significant links are coloured grey else remain white Festuca rupicola 208425 0 01400 Po _paraste ost prot vae __ probe vale 0 01400 x 359 Entyloma_dactylidis_63287 Poa_annua_93036 0 01200 0 01200 36
21. holds if we consider non homologous loci for use in supertree reconstruction let alone the current debate about whether and how to infer supertrees However this may not be true for all taxonomic groups Since there most likely is a trade off between the number of taxa and the topological resolution available as input for ParaFit the user has to decide whether a certain ParaFit analysis based on taxonomic data is worth conducting or not It is therefore necessary to closely examine the number of resolved associations compared to those of a study of the same taxonomic group but based on character data as well as the resolution of the host or parasite taxonomy trees as described below Computation of the broken stick distribution BSD for the set of associations The user selects an AT which results in a new window displaying the BSD see figure 4 It basically consists of two parts the first part shows the BSD for the parasites indicated by a P in the first column of each line the second part the BSD for the hosts a H as identifier The further columns show the rank of the taxon its name its absolute and relative frequency of occurrence within the associations and its expected frequency according to the BSD The last column shows whether the real relative frequency is larger than the expected one By holding the CTRL button and using the left mouse button the user can highlight multiple entries within the list Each entry repre
22. my dump files is steadily increasing potentially increasing both the number of resolved associations and the topological resolution in the NCBI based cophylogenetic analyses By checking this option CopyCat downloads the latest NCBI taxonomy dump file and places it in the appropriate directory 12 View View content of working directory This shows a view of the current working directory and its content Options Enable Strict Filtering of Association Table By default the filter process scans an association table and removes all associations lines which do not fulfill one of the following criteria Both parasite and host label have to exist in the NCBI taxonomy Should not be blacklisted If the user has provided an association table containing custom taxon labels such as Patient234 instead of Homo sapiens the program would remove this line from the new associations table due to the first condition have to exist in the NCBI taxonomy Even though this condition can be relaxed by unchecking the option Enable Strict Filtering of Association Table in the menu bar Use Equal Branch Length 1 for tree2dist Conversion If you have specified a tree file in CopyCat s second tab you might want to have topological distances in the distance matrix resulting from that tree By checking this option branch lengths are set to 1 By default this option is not checked In that case a
23. nthracoidea_arenaria_265863 Carex_brizoides_240677 6 Anthracoidea_arenaria_265863 Carex_ligerica_240664 7j Anthracoidea_arenaria_265863 Carex_ovalis_140840 8 Anthracoidea_arenaria_265863 Carex_praecox_240661 9 Anthracoidea_aspera_251614 Carex_appropinquata_240679 10 Anthracoidea_aspera_251614 Carex_chordorrhiza_240692 11 Anthracoidea_aspera_251614 Carex_diandra_140803 Anthracoidea_bigelowii_265841 Carex_bigelowii_241200 Anthracoidea_buxbaumii_265847 Carex_buxbaumii_241202 Anthracoidea buxbaumii 26584 Carex hartmanii Anthracoidea_capillaris_265856 Carex_capillaris_241203 Anthracoidea_caricis_265852 Carex_montana_140837 Anthracoidea_caricis_265852 Carex_pilulifera_140848 Anthracoidea_caricis albae_265855 Carex_alba_241195 J Anthracoidea_buxbaumii_265847 Carex hartmanii apply changes dump results to working directory Illustration 8 Copycat s resolving step Associations highlighted in light grey have a valid taxonomy ID for both the parasite and the host By clicking the dump results to working directory button the resolved association table is written to the working directory a subset of the original table The file name contains a _resolved txt suffix 3 Generation of the parasite host taxa lists For the Parafit LDB02 analysis of this association data we first need to draw two lists from the resolved association table one containing all parasite taxa and another one containing all host taxa This is achieved by
24. see figure 3 tries to assign a NCBI taxonomy ID for each organism name The set of IDs is necessary for inferring the NCBI host or parasite tree The user specifies a file unresolved association table containing one parasite and one host name per line tab separated Then CopyCat tries to retrieve a NCBI taxonomy ID equivalent to each entry e g the taxonomy ID 9606 for the host name Homo sapiens These results can be watched by clicking the show edit results button This button is enabled once the results are complete Figure 3 shows an example of this result window Anthracoidea_arenaria_265863 Anthracoidea_arenaria_265863 Anthracoidea_arenaria_265863 Anthracoidea_arenaria_265863 Anthracoidea_arenaria_265863 Anthracoidea_aspera_251614 Anthracoidea_aspera_251614 Anthracoidea_aspera_251614 Anthracoidea_bigelowii_265841 Anthracoidea_buxbaumii_265847 Anthracoidea_capillaris Anthracoidea_caricis_265852 Anthracoidea_caricis_265852 Anthracoidea_caricis albae_265855 Carex_arenaria_234466 Carex_brizoides_240677 Carex_ligerica_240664 Carex_ovalis_140840 Carex_praecox_240661 Carex_appropinquata_240679 Carex_chordorrhiza_240692 Carex_diandra_140803 Carex_bigelowii_241200 Carex_buxbaumii_241202 Carex hartmanii Carex_capillaris_241203 Carex_montana_140837 Carex_pilulifera_140848 Carex_alba_241195 Anthracoidea_buxbaumii_265847 Carex hartmanii lt A apply changes dump results to working directory h x Figur
25. selecting apply settings to association list Here the specified association list can be filtered in regard to certain criteria and as a side effect the parasite host taxa lists are written to the working directory Naturally the user is not obliged to select certain filter criteria but can simply choose the leave associations in their current state option Consequently the specified association table stays untouched In this tutorial we make use of the latter and issue the association table gained in the previous section This operation will take a moment Finally the following two files appear in the working directory hosts filtered using option 0 txt parasites filtered using option _0 txt The 0 indicates that we selected the leave associations in their current state option 4 Creation of a host distance matrix and a parasite distance matrix The taxa 16 lists from the previous step are now being used for the creation of the respective NCBI host tree and NCBI parasite tree Once this is done the respective distance matrices are generated We switch to Copycat s second tab and select the distance matrix from host taxa list option together with the hosts filtered_using_option_0 txt file This results in the call of the ParafitWrapper The wrapper will now try to create the denoted host distance matrix You might want to follow the process of the wrapper by reading the lines in the message window that are marked
26. sents a parasite or host which is then marked for removal from the association table Often the entries in the list are highlighted by means of an alternating pattern of dark and light grey Several lines sharing the same colour scheme correspond to the same tied rank and therefore have to be treated equally broken stick results P parasite H host blocks of tied ranks are highlighted Microbotryum_violaceum_5272 0 04456 0 02148 Jamesdicksonia_dactylidis_63287 0 03539 0 01985 Ustilago_bullata_117172 0 02752 0 01863 Urocystis_ranunculi_63394 0 02752 0 01766 Thecaphora_saponariae_72562 0 0249 0 01684 Tranzscheliella_hypodytes_349358 0 02359 0 01615 Entyloma_hieracii_189602 0 02359 0 01554 Entyloma_ranunculi repentis_189607 0 01966 0 01499 Entyloma_microsporum_62642 0 01704 0 01451 Ustilago_avenae_120650 0 01573 0 01406 Ustilago_tritici_117174 0 01442 0 01366 VVVVDDV DD VDD Anthracoidea_karii_265844 0 0118 0 01261 Tilletia_laevis_157183 0 0118 0 0123 Ustilago_hordei_120017 0 0118 0 01202 Entorrhiza_casparyana_63375 0 0118 0 01174 Microbotryum_duriaeanum_162322 0 00917 0 01037 Anthracoidea_fischeri_251615 0 00917 0 01018 Figure 4 CopyCat s representation of the Broken Stick Distribution N B The broken stick distribution e g Legendre and Legendre 1998 p 244 is a standard null model of community structure in ecology It can be used to predict species relative abundances but may also be used with other kind of data s
27. uch as e g eigenvectors Legendre and Legendre 1998 p 410 Species the relative frequency of which is larger than the corresponding broken stick value occur more frequently than expected by chance We have included the BSD here since it may be used to detect host or parasite species which are represented in significantly more associations than others This is not to say that ParaFit is unable to deal with widespread parasites on the contrary these are treated more consistently in ParaFit than in other 6 cophylogeny programs Legendre et al 2002 However a list of associations derived from literature data may for instance include many more associations from host species which are medically or economically important and thus have been studied more intensively than their less important relatives If the BSD detects species which are represented in a particularly large number of associations the user may wish to conduct ParaFit runs both before and after exclusion of these taxa In case such taxa display a cophylogenetic behaviour strongly deviating from that of other taxa i e significant vs insignificant associations or vice versa presence or absence of these highly frequent taxa may considerably influence global significance Even though ParaFit s results will according to Legendre et al 2002 always be the correct ones given the correctness of the associations the user may be interested in the impact of such extremely widespr
28. y formatted input data amp CopyCat The cophylogenetic analysis tool this session has ID 8 4 File View Options Setup Help Preprocessing of Input Data iC Evaluation of Parafit Results m step 1 select parameters for Parafit selectthe number of permutations used in Parafit 999 X correction method used in DistPCoA 3 no correction step 2 select association file selectfile r step 3 create host distance matrix create distance matrix from host taxa list C create distance matrix from hosttree C use already existing distance matrix select host taxa list and compute matrix r step 4 create parasite distance matrix create distance matrix from parasite taxa list C create distance matrix from parasite tree C use already existing distance matrix select parasite taxa list and compute matrix step 5 input data validation and start of analysis validate the specified data peparedatatonremote anelysic elcome to CopyCat The cophylogenetic analysis tool INFO The operating system is WINDOWS 2000 INFO The working directory is E Eclipse workspace Copycat default wdir ID_8 INFO In order to change the configuration check show setup menu at next program start in the Setup menu Y Figure 5 The second tab of CopyCat This tab is divided into 4 steps Step 1 The number of permutations per row of the asso
29. yloma_dactylidis_63287 Agrostis_canina_218142 0 02300 0 02300 375 Entyloma_dactylidis_63287 Agrostis_capillaris_204232 0 02400 0 02400 376 Entyloma_dactylidis_63287 Agrostis_stolonifera_63632 0 02100 0 02100 377 Entyloma_dactylidis_63287 Agrostis_alba_29659 0 02600 0 02600 378 Entyloma_dactylidis_63287 Phleum_pratense_15957 0 02100 0 02100 379 Entyloma_dactylidis_63287 Alopecurus_pratensis_15304 0 02900 0 02900 380 Entyloma_dactylidis_63287 Holcus_lanatus_29679 0 02300 0 02300 381 Entyloma_dactylidis_63287 Avenula_pubescens_87471 0 02300 0 02300 382 Entyloma_ossifragi_272710 Narthecium_ossifragum_114204 0 70600 0 70600 383 Rhamphospora_nymphaeae_62645 Nymphaea_alba_34301 0 30900 0 30900 384 Doassinga_callitrichis_62640 Callitriche_stagnalis_119595 0 01500 0 01500 x the pre defined significance value 0 02 apply Overall cophylogenetic structure is highly significant 0 00100 lt 0 02 sig val 538 links out of 634 are significant dump information to file Illustration 9 Copycat s representation of the Parafit results References Colless D H 1982 Review of Phylogenetics The Theory and Practice of Phylogenetic Systematics by E O Wiley Syst Zool 31 100 104 Farris J S 1967 The meaning of relationship and taxonomic procedure Systematic Zoology 16 44 51 Legendre P and Anderson M J 1998a Distance based redundancy analysis testing multi species responses in multi factorial ecological experiments Ecological Mono
Download Pdf Manuals
Related Search