Thank you everyone for your answers and suggestions. Here are the replies I got. Best; Omur ****** For topology reconstruction, Takezaki & Nei (1996) suggest that in both the Infinite Allele Model (IAM) and Stepwise Mutation Model (SMM), Cavalli-Sforza and Edwards (1967) chord distance and Nei et al. s (1983) genetic distance are preferably chosen, while for estimating branch lengths, (dľ)˛ may yield better results. I attached Takezaki & Nei's paper to this email for you. With the fitch-margoliash algorithm, implemented in FITCH in PHYLIP, you can construct such a hybrid tree, using e.g. Nei's genetic distance for topology and delte mu 2 for branch lengths. I used to calculate delta mu 2 values in MSA. Dieter.Anseeuw@kuleuven-kortrijk.be Unless your individuals are all of different species they don't have a "phylogenetic relationsip". Instead each locus has a different coalescent tree. By forcing them all to have the same tree you are distorting the relationships. You should look into programs like Migrate or IM. If you're thinking in terms of phylogenies of individuals you should re-think the matter. J.F. ---- Joe Felsenstein joe@gs.washington.edu Department of Genome Sciences and Department of Biology, University of Washington, Box 357730, Seattle, WA 98195-7730 USA Depending on your sample size, you many be able to use more appropriate methods for determining the relationships between populations. Phylogenetic methods were developed for reciprocally monophyletic species and should really not be applied to populations that exchange genes. There are now some better methods to determine the relationships among populations. Two packages that I have found quite useful are Structure (http://pritch.bsd.uchicago.edu/ structure.html) and Population Graph (http://dyerlab.bio.vcu.edu/ wiki/index.php/Software). David B Lowry delta mu square values can (as far as I know) only be calculated with RSTCALC. There are however other ways to conduct a phylogenetic analysis based on allele frequency data, e.g. Nei's genetic distance. I am attaching a protocol I wrote for doing so. It involves the following softwares "Convert", "Phylip" and "Treeview". All are available for free. ______ Dr. Susanne Hauswaldt Unit of Evolutionary Biology and Systematics Department of Biological Sciences University of Potsdam here's a list of programs that compute genetic distances from microsatellite data as well as a distance tree. Some even boostrap across loci. The best known program for microsatellite distances is MICROSAT: http://hpgl.stanford.edu/projects/microsat Microsatellite Analyzer or MSA (does all kinds of distances, among other things): http://i122server.vu-wien.ac.at/MSA/info.html/ MSA_info.html#distance Populations: http://bioinformatics.org/project/?group_id=84 POPTREE (just to make the NJ tree): http://www.bio.psu.edu/People/ Faculty/Nei/Lab/software.htm SHAREDST (across individuals): http://www2.biology.ualberta.ca/ jbrzusto/sharedst.php GeneDist (across populations): http://www2.biology.ualberta.ca/ jbrzusto/GeneDist.php Sergios-Orestis Kolokotronis Dept. of Ecology, Evolution, and Environmental Biology Columbia University sk2059@columbia.edu Try the AFC plots in the GENETIX (Belkir et al.) software. The plots are not based on delta-mu ^2, they partition variance instead. Yet, each individual will be represented in a multi-vector space. scott _________ Scott M. Blankenship Ph. D. Biologist/Geneticist Washington Department Fish and Wildlife blanksmb@dfw.wa.gov ******** In the analysis of microsatellite loci, if the number of loci examined is small, Da or chord distance should work better than delta myu square to estimate the phylogenetic relationship of populations. Previously, we have studied the probability of obtaining a correct tree by using Da, chord distance, delta myu square, and some other distances for microsatellite loci and showed that Da and chord distance work better than delta myu square unless hundreds of loci are used (Takezaki and Nei 1996). This is because even though the value of delta myu square is expected to increase linearly with time under the stepwise mutation model that the microsatellite loci roughly follow, the sampling error of the delta myu square is used. More explanations about genetic distance and phylogenetic tree are found in chapter 13 of the book of Nei and Kumar (2000). In this chapter, the example of the phylogenetic trees constructed from 25 microsatellite loci of human populations is shown. The tree constructed using Da distance looks reasonable. However, even with the use of 25 loci, in the tree made with delta myu square populations from the same geographic region does not form clusters. Da distance can be computed and used for constructing the phylogenetic tree with neighbor-joining method or UPGMA by a computer program DISPAN and POPTREE. POPTREE can also compute delta myu square for constructing phylogenetic trees. DISPAN and POPTREE are available at the following webpage. http://www.bio.psu.edu/People/Faculty/Nei/Lab/software.htm References Takezaki, N. and M. Nei. 1996. Genetics 144: 389-399. Nei, M. and S. Kumar. 2000. Molecular Evolution and Phylogenetics. Oxford University Press, New York. Best regards, -------- Naoko Takezaki Information Technology Center, Kagawa University 1750-1 Ikenobe, Mikicho, Kitagun, Kagawa 761-0793 Japan TEL/FAX 087-891-2419 email: takezaki@med.kagawa-u.ac.jp ****************** Omur Kayikci Associate in Research Magwene Lab DCMB Group, LSRC B232 Duke University Omur Kayikci