Dear EvolDir members, thanks to all who responded to my request about software for coalescence simulations. I have got many useful answers and post them to the list. Most of you suggested either using SimCoal / SimCoal2 or generating genealogies using MS and then and then using a sequence simulation program, like SeqGen, to generate sequences on those genealogies. It seems that simulating selection together with complex demographies under the coalescence is not yet possible and/or feasible. One comprehensive reply listed a number of software packages: MS http://home.uchicago.edu/~rhudson1/source/mksamples.html SIMCOAL/SIMCOAL2 http://cmpg.unibe.ch/software/simcoal2/ Serial SIMCOAL http://www.stanford.edu/group/hadlylab/ssc.html SPLATCHE http://cmpg.unibe.ch/software/splatche/ Coasim http://www.daimi.au.dk/~mailund/CoaSim/ TREEVOLVE http://evolve.zoo.ox.ac.uk/software.html?name=Treevolve SEQ-GEN http://evolve.zoo.ox.ac.uk/software.html?id=seqgen EASYPOP http://www.unil.ch/izea/softwares/easypop.html BOTTLESIM http://chkuo.name/software/BottleSim.html SIMUPOP http://bp6.stat.rice.edu:8080/simuPOP/ FPG http://lifesci.rutgers.edu/~heylab/HeylabSoftware.htm#FPG Other replies and the original request are listed below. thanks again, Asta ** Can anybody suggest a software package for coalescence simulations under different population demographic models (subdivision, stable population size, bottlenecks) and different background mutation rates? Ideally there should also be a possibility to introduce selection, different mutation models, and I’d prefer the results (a random sample of sequences from the simulated populations) to be available in .nex or some other popular format. Answers: I do not now of any software that do all the stuff that you want would like to recommend Hudsons MS http://home.uchicago.edu/~rhudson1/ source/mksamples.html and if you want to include selection try SelSim http://www.stats.ox.ac.uk/~spencer/SelSim/Controlfile.html. At least ms is used extensively and once you get started fairly easy to use. ** try SimCoal or SimCoal2 (http://cmpg.unibe.ch/software/simcoal2/) ** have a look at serialsimcoal ** Hi, you have a number of options. Probably the most efficient neutral simulator would be Richard Hudson's ms (http://home.uchicago.edu/ ~rhudson1/source/mksamples.html), the results of which can be piped into Andrew Rambaut's SeqGen (http://evolve.zoo.ox.ac.uk/ software.html?name=Seq-Gen) to generate sequences under varying mutation models. For selection, have a look at Spencer & Coop's SelSim (http://www.stats.ox.ac.uk/~spencer/SelSim/Controlfile.html). All three are command line unix-type programs which are easy to incorporate into larger shell or perl scripts if necessary. Another option for selection simulations would be HYPHY (http:// www.hyphy.org/) which provides a framework for phylogenetic hypothesis testing. Selection can be simulated along phylogenetic trees from predefined roots under given models. Again unix-style batch processing is possible. ** The most popular program for coalescent simulation is ms (Hudson, Bioinformatics 2002). you can download at http://home.uchicago.edu/~rhudson1/ Additionally, another programs have been constructed. I designed a program named mlcoalsim (Evolutionary Bioinformatics 2007) based on ms that construct sample sequences in fasta format and also calculates several neutrality tests. you can download at http://www.ub.edu/softevol/mlcoalsim ** I have a program which should allow you to do this (except for implementing selection) at ueg.ulb.ac.be/treesSifter/ I however suggest you take a close look at SIMCOAL 2 (cmpg.unibe.ch/ software/simcoal2/) and at Richard Hudson's MS program (home.uchicago.edu/~rhudson1/source/mksamples.html). ** See software developed by Sebastian Ramos Onsins in Barcelona http://www.ub.es/softevol/mlcoalsim/ ** We tend to use Simcoal2 http://cmpg.unibe.ch/software/simcoal2/ which most probably has everything you want with the exception of different mutation models for sequences and selection. We have done a easy to use graphical interface for it: http://popgen.eu/soft/m4s2/ Regarding coalescent selection simulation there is selsim, but I think it is single population constant size only. Regarding different mutation models for sequences, I don't know of any simulator that does it, if you find one, I would really appreciate if you could inform me... ** I found EASYPOP quite useful. F. Balloux EASYPOP (Version 1.7): A Computer Program for Population Genetics Simulations J Hered 92: 301-302. http://www.unil.ch/dee/page6759.html ** I suspect your request might be overly optimistic. The best software available for coalescent simulation is still probably Richard Hudson's 'ms'. This programme lets you simulate effectively any demography you wish, including subdivision, constant population sizes, bottlenecks, etc under user-defined thetas ("mutation rates"). The code is available here: http://home.uchicago.edu/~rhudson1/source/mksamples.html If you are interested in different mutation models, you can feed the output from 'ms' into 'seq-gen'. However, this does not appear to be a standard approach in coalescent modelling -- most researchers use a more standard binary infinite-sites output instead. As far as I'm aware, there is no software (or indeed method) available to simulate selection on top of complex demographies under the coalescent. If you discover that I'm mistaken on this point, please let me know! Finally, most coalescent output is software-specific, and I suspect that few coalescent modellers would choose to use nexus format or any other DNA-like file formats anyway (e.g. fasta, prettybase). In any case, you will quickly find that any research beyond very simple coalescent modelling requires some programming skills -- at the very least in languages like Perl or R. If you really need simulation output in nexus format, you will probably have to write the code yourself. You might also like to take a quick look at my website at www.u.arizona.edu/~mpcox/ ** I think there are a few packages available to you - check out: coaSim: http://www.biomedcentral.com/1471-2105/6/252#B2 selSim: http://bioinformatics.oxfordjournals.org/cgi/content/abstract/ bth417?ijkeyzRFUSCPcvX2&keytype=ref msHot: http://bioinformatics.oxfordjournals.org/cgi/content/abstract/ btl622v1 For phylogeography - David Posada has implemented templeton's methods http://darwin.uvigo.es/ ** I think you could use Serial Simcoal-http://iod.ucsd.edu/simplex/ssc/ (the link is indicated)for all the parameters indicated-different population demographic models (subdivision, stable population size, bottlenecks) and different background mutation rates. ** One option is to use Hudson's ms to simulate genealogies (ms will do subdivision, bottlenecks, growth, selection, recombination, etc...), and then use a sequence simulation program - e.g., seq-gen - to generate sequences on those genealogies. ** Dick Hudson's MS program ( http://home.uchicago.edu/~rhudson1/source/ mksamples.html ) piping to SeqGen ( http://evolve.zoo.ox.ac.uk/ software.html?id=seqgen ) . This will do most of what you want, except for non-neutral evolution. ** [Another relevant message that was posted to EvolDir recently] For all users interested in using the coalescent but consider that most programs are hard to use, we have developed a easy to use GUI for coalescent simulations (currently using simcoal2) called modeler4simcoal2 (m4s2) For anyone interested in an easy way to use the coalescent, you can visit: http://popgen.eu/soft/m4s2/ or: http://bioinformatics.oxfordjournals.org/cgi/content/abstract/btm243v1 *** Asta Audzijonyte Finnish Museum of Natural History POB 26 (Teollisuuskatu 23) FI-00014 University of Helsinki www.fmnh.helsinki.fi/english/zoology/invertebrates/research/molecular/asta.htm audzijon@mappi.helsinki.fi