I'm constructing a phylogenetic tree using microsatellite sequences in MEGA11. I have aligned the sequences using MUSCLE and now I can see a lot of "xxx" and gaps in many of the sequences. Kindly suggest whether I should remove the entire sequence or should just remove the part of the sequence. For better understanding provide an answer with an explanation.
Thank you in advance.
Tricky question, but at first, you may think whether the phylogenetic method you intend to use will assume gaps as mutational steps or just will ignore them as missing data (I believe this is the default in MEGA).
Nonetheless, you must bear in mind that an alignment is a homology hypothesis for each site. I highlighted that because MUSCLE (and other algorithms) will result in several very likely alternative alignments for challenging sequences such as microsatellite loci. I recommend you take a look at Satè as an alternative algorithm. Analyze your alignment carefully before proceeding to phylogeny estimations.
I believe it is advisable to avoid using gaps as mutational steps unless you are pretty sure the alignment is reliable. You must think about whether you can firmly believe that your gaps are mutational steps. If so, use BEAST with bModelTest as model estimation or mrBayes with a tested evolutionary model combining your nucleotide matrix with a binary matrix corresponding to your gaps (cut it out and change any nucleotide to 1 and gaps to 0).
Otherwise, you can just run your phylogenetic estimation with any algorithm you may find reliable (distances might be good, parsimony might be also good...) Look at the MEGA handbook and check it ou whether parsimony accepts gaps as mutational steps. Also, pay attention to deletion modes (pairwise, complete deletion, or missing percentage criteria).
Sorry for not being more specific, but I believe it is better you be aware rather than follow instructions.
Good luck!
Thank you Wilson Frantine-Silva for throwing light on parameters that I should consider before proceeding to the phylogeny tree. It was very helpful. To give you more info, I'm not considering the mutation here since I just wanted to construct a phylogeny tree to understand how related are their phenotypes to each other.
Great then. So I would tell you to go forward with your phylogeny once you are confident in your alignment. Again, deletion modes might play a key role on your estimations.
Best
- Badges
- Science topic