Entering edit mode
6.4 years ago
a.m.dekker
•
0
Dear everybody!
I have a question regarding building phylogenetic trees. I am working with pooled (n = ~ 500) PACBio sequencing material with reads of ~ 4kb length. I am not very familiar with the current methods for building trees, so was wondering whether someone could help me out! I tried clustalw2 to draw an 'instant' tree and is really fast, but does not show very good results so far. I also tried RAxML by converting a vcf to a phylip, but I believe this does not take indels that well into account.
Thank you so much in advance, Annabel
How/why did you decide that? Just by visual inspection? What kind of sequences are you working with (amplicons)? Are they homologous (enough) to actually be able to infer a phylogenetic relationship?
Thank you for your question, I should be more clear. Yes, basically by visual inspection. I am indeed working with amplicons and trying to visualize variation in one gene between different samples of a plant species (I thus assume homology, yes). I decided that my tree is sloppy, because from some/many of the samples i do know the gene content and thus know that they should be clustered together.
If you have access to SMRTanalysis then you may be able to use this.
Thank you for your suggestion! Your link shows me how to preprocess my pacbio data for clustering raw reads for creating consensus reads (which is already done in my case), is that correct? I am looking for an algorithm suitable to cluster these consensus long amplicons for sample analysis.