Constructing Gene Trees From Partial Sequences

Question

[hicsuntdrac0nis]

I have a bunch of partial sequences from a Metagenomic Database . They are all 16s partial sequences . The sequences have a lot of overlap and I'm trying to construct a tree of the similarities between each sequence . Any idea on where to start with this task ? I was thinking of using PAUP but I can't get it to work on Macintosh .

Answer 1

[Joseph Hughes]

You could start with a Nj or UPGMA tree in Clustal. This will provide you with a rough guide tree to start with. You could then test which model would be best to use using ModelTest and then move on to using RAxML and MrBayes for some reliable phylogenetic inferences. The standard way of treating both gaps and missing data in likelihood-based phylogeny programs is to treat them as missing data. However, the accuracy of the reconstruction will be influenced by the amount and distribution of missing data resulting in tree reconstruction artefacts.

If the plan is to place the sequences on a known reference tree, you might want to look into pplacer or PaPaRa.