How to tell whether an MSA result is good and how to set parameters?

Question

[wuqingkuangsha]

Hi everyone, I am a student currently studying multiple sequences alignment and phylogenetics. I am just confusing about how to tell if an MSA result is good. I always run an alignment tool using the default settings. But I also read that parameters can strongly affect the result. I agree with that but I don't know how to tell if I really need to change some parameters.

Answer 1

[dago]

I think there are different approaches to asses the alignments:

A statistical score for assessing the quality of multiple sequence alignments

Automatic assessment of alignment quality

However, I personally never really dug too much into it. For two main reasons:

1. I am often familiar with the protein/gene I am working with and by looking at the alignments I can see whether there are enough conserved regions (someone calls them blocks)
2. sometimes alignments are simply too big. For this, there might be more "brutal" force approaches as trimming "noisy regions" using e.g. Trimal or Gblocks. Another option would be to use Zorro that give scores of conservation to the different positions.

This question is quite interesting though and I am curious to read the approaches of the more expert users.