I want to construct a phylogenetic tree using SNP rather than amino acid or DNA seq. Is it also possible to make a tree using snp through diverge phylogeny?
Yes of course, there are many software that construct phylogenetic tree (called neighbor joining) using genotyping data (genetic difference between individuals or species by molecular markers: microsatellites, SNPs...). I hope I got your question right and I answered it. If not, please explain more.
There are a lot of different approaches how to use SNPs in phylogeny now. It's easy because these events can be recognized as nucleotide substitutions and very useful for many reasons i.e. genotyping and others. Most of modern algorythms in phylogeny allow to do this with no problem. But note that in case in you planed to work only with aminoacid sequences you can observe only significant canges (which caused aminoacid change -> changes 1-3 nuleotides in a triplet), but can't detect variable positions of DNA (that not caused any changes in aminoacid sequence). But also many things will depend on you r ains and kind of data you use in your approach.
Some papers on this topic which can be useful:
http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1003215
http://link.springer.com/chapter/10.1007/978-0-387-36011-9_6#page-1
http://jb.asm.org/content/188/2/759.short
http://journals.plos.org/plosgenetics/article?id=10.1371/journal.pgen.1002451
You can use RAxML with the ASC parameter for variable sites. If you don't tell an ML algorithm that you have only variable sites then the likelihood is calculated incorrectly. A better suggestion for smaller datasets is SNAPP in the BEAST package, which considers the coalescence of each site independently.
You might also want to take a look at the SNPhylo pipeline; this can handle very large numbers of SNPs:
http://chibba.pgml.uga.edu/snphylo/
If your data is from different individuals or populations you can use the simple one-click software VCF2POPTTREE (http://sankarsubramanian.net/dat/index.html) that builds an UPGMA/NJ tree in seconds.
Konoutan Médard Kafoutchoni this is very possible.
From the SNP data, you can generate a dosage format using plink. from the dosage format, you can generate the IBS dissimilarity matrice or Jaccard matrix. From there you can use library ape to construct your phylogeny tree.
- Badges
- Science topic
- Similar topics
- Filing