1
$\begingroup$

I have an MSA (protein sequence) and I have used various programmes (Clustal, Aliview, MEGA11 etc) to align. However, in all programmes I get many gaps which is not ideal as I am trying to construct a tree from this alignment.

My question is are there any tools to delete gaps in alignments?

Your help would be greatly appreciated.

$\endgroup$

4 Answers 4

2
$\begingroup$

You may also find useful the BMGE tool (stands for Block Mapping and Gathering with Entropy). BMGE selects regions in a multiple sequence alignment that are suitable for phylogenetic inference by computing their entropy-like score. This allows to distinguish among biologically expected and unexpected variability for each aligned character. BMGE also implements different types of trimming (for instance, deletes gaps) to minimize phylogeny reconstruction artifacts.

Available from here and also installable from Bioconda.

$\endgroup$
2
$\begingroup$

If you want an ungapped alignment, you could do an ungapped alignment (the older read aligners were ungapped, e.g. bowtie). However, I suggest that you don't want an ungapped alignment, you just want a stricter one, here's a few options...

  • Remove/trim poorly aligning sequences. You could try GUIDANCE, which has an explicit model or GBLOCKS for a more simple filtering criteria
  • Tune your alignment parameters for ClustalOmega/MAFFT etc. to make it stricter
  • Finally, consider that your sequences must be homologous otherwise you'll trim away sequences until you have a random alignment that appears good
$\endgroup$
1
$\begingroup$

Gblocks is a tool for removing gaps/poorly conserved positions from alignments. Available here. Here is a webserver that runs it for you.

It's a bit old, possibly there are others at this point.

$\endgroup$
1
$\begingroup$

This new tool, CIAlign, talks about some automated removal of gaps in multiple sequence alignments

https://twitter.com/thekatybrown1/status/1306215132670889984

https://github.com/KatyBrown/CIAlign

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.