Is there a way to remove redundant amino acid sequences from a fasta file but also output all the redundant accession numbers in groups, just like mothur's unique.seqs command (which unfortunately only works on nucleic acids data).
the accession number output should look like this (or similar):
G9SS7BA01AM9A3 G9SS7BA01AM9A3,G9SS7BA01EMTMV,G9SS7BA01CYG40,G9SS7BA01AWI8Z,G9SS7BA01AFVJC,G9SS7BA01BCZCD,G9SS7BA01DBN7B,G9SS7BA01CZ7GO,G9SS7BA01C05FB
G9SS7BA01EAKDX G9SS7BA01EAKDX,G9SS7BA01B1MNY
G9SS7BA01C2SRQ G9SS7BA01C2SRQ,G9SS7BA01AK1UJ,G9SS7BA01BLVCZ,G9SS7BA01ARMFA
G9SS7BA01BQ5UG G9SS7BA01BQ5UG,G9SS7BA01BZ9XF
G9SS7BA01BD4F9 G9SS7BA01BD4F9
Where each row is a group of identical seqs and the first column is the one kept in the 'uniques' file.
USEARCH only outputs a file with the unique seqs.
•
written
by
angel.roey •
3 answers
You can use CD-HIT (http://weizhong-lab.ucsd.edu/cd-hit/) and parse the resulting cluster file into tab delimited format.
The only problem you might face with CD-HIT is that if your sequence IDs are really long, the cluster output file will shorten the name automatically. You might have rename your fasta files first to a shorter name and then remap the names back afterwards.
Nice. I didn't know that. I guess I should really go through the options.
Thanks! Any possibility of influencing what is being written to the CLSTR file? e.g. only include seqs with redundancy, change format etc.
Doesn't matter, USEARCH does all I need.
Thanks! It does. And in a nicer format than cd-hit
Here is my free program on Github Sequence database curator (https://github.com/Eslam-Samir-Ragab/Sequence-database-curator)
It is a very fast program and it can deal with:
- Nucleotide sequences
- Protein sequences
It can work under Operating systems:
- Windows
- Mac
- Linux
It also works for:
- Fasta format
- Fastq format
Best Regards
Log in to answer this question.