search

FePhyFoFum / phyx

phylogenetics tools for linux (and other mostly posix compliant) computers
blackrim.org
GNU General Public License v3.0
115 stars 17 forks source link

Using pxclsq to trim codon alignments #90

Open flopezo opened 6 years ago

flopezo commented 6 years ago

Hello,

I was wondering if maybe there is a way to use pxclsq to remove codons instead of single sites; something similar to what this script does.

Thank you.

josephwb commented 6 years ago

Definitely!

josephwb commented 5 years ago

Gah this clearly fell through the cracks. Didjya want to do this @jfwalker (since you wrote it)? If not, I can do it.

jfwalker commented 5 years ago

@josephwb so I don't think it should matter whether it is nucleotide or codon. For codon it will likely have been aligned by an Amino acid or an algorithm that assumes three's such as prank. This means when you do the column occupancy all bases in the codon should either be present or absent together so if it's removing position one it will have to remove position 2 and 3, unless somehow this was aligned with certain bases of the codon missing. Running on regular nucleotide should be fine for codons. Sorry completely forgot to respond to this before!

josephwb commented 5 years ago

I think the idea is to remove the stop codon, even if it is present in all the sequences.

josephwb commented 4 years ago

Sorry this has taken forever, but it has been implemented as of 0154ac6d8f70dc26358c3c0e47b78f01df5bf9c9. Still need to add the bit about removing the stop codon. Will close this after merging to master.