Multiple sequence alignement MSA editing

Question

[TEman]

I want to remove all rare insertions (when it occurs in less than 5% of the sequences) in a multiple sequence alignment file (clustal .aln) with 699 sequences.

That is, I have a MSA with many columns containing only one or two insertions while the rest of the sequences are blank "-". It is by far too much to do manually.

Any suggestions how to do this?

[Joe]

Do you specifically want to do this in R?

If you use BioPython, you can create an ungapped concensus sequence with a threshold for inclusion of a particular residue in a column.