Dear All, I have posted the following question on evoldir: Hi to everybody, I have a question about the Tajima's D analysis implemented in DNAsp: in some papers (for instance Wheat et al. 2006, Mes et al. 2006) I found that the authors use the tajima's D analysis considering only the synonymous sites, but I don't really find the way to do it. Is there someone who knows how to do it? All answers would be very helpful. I would like to thank all that answered me, it was really helpful. Here are all the responses I received, hope it will be helpful for someone else! Again, thanks to everybody! ##### Dear Erica, please note that in my methods section (Wheat et al. 2006) I state that I used two programs for DNA summary statistic calculations. DNAsp did not do syn site Taj D calculations then, and I don't think it does it now- which is a bit of a pain. However, the other program, Proseq, does do syn site Taj D calculations- but be careful, the new program version is a bit tricky to use. I strongly recommend estimating your TajD calculations using more than one program if you are going to use the point and click GUI programs. This will ensure you are getting the numbers you really want. Cheers, Chris ######### This really is surprisingly difficult to do, and the work-around seems to spread slowly by word of mouth! (1) Export the synonymous polymorphic sites as a new file: generate>polymorphic/variable sites datafile>only synonymous (2) Open the new file, and do the Taj D analysis on that [If you are then using coalescent simulations with intermediate levels of recombination, remember to use the rate for the whole lenght of the gene, not just the subset you've generated!] Good luck! Darren ########### Hello Erica, I believe that the way to do it is to created an input file with only synonymous SNPs. In this case you will calculate TD only considering synonymous sites. Best, Luis ########## Hi, Under the Data menu, go to Remove Positions, then select which positions to include/exclude. Every analysis you do after that will only be performed on the positions you have decided to keep. I hope that helps. Brian ######## Dear Erica: We've done it a lot. I find that looking at synonymous and nonsynonymous sites separately with Tajima's D-like statistics is very revealing. I actually use the ratio of D to the absolute value of its theoretical minimum. Thie latter statistic (Q) has the advantage of not being dependent on sample size (unlike D) and therefore comparable across data sets. We have done this mostly using custom-made software, some of which we are gettiong into a user-friendly form for publication in the not too distant future. Austin Hughes Some references: Hughes, A.L. 2005. Evidence for abundant slightly deleterious polymorphisms in bacterial populations. Genetics 169: 553-558. Hughes, A.L., and Hughes, M.A. 2007. More effective purifying selection in RNA viruses than in DNA viruses. Gene 404: 117-125. Hughes, A.L., and Hughes, M.A.K. 2007. Coding sequence polymorphism in avian mitochondrial genomes reflects population histories. Mol. Ecol. 16: 1369-1376. Hughes, A.L., and Piontkivska, H. 2008. Nucleotide sequence polymorphism in circoviruses. Infection, Genetics, and Evolution 8:130-138. Dr. Erica Bortolotto Biology Dept University Of Padova Via G. Colombo 3 I-35100 Padova Italy e-mail bortolotto@unisi.it Tel 0039 049 8276222 Fax 0039 049 8276209 Erica Bortolotto