I am having unaligned multiple sequence FASTA file (don't want to align) with ambiguous characters: RYWSMKHBVDNS
I want to remove/delete those characters...
1) I tried with mEMBOSS
Readme: http://emboss.sourceforge.net/apps/release/6.6/emboss/apps/trimseq.html#input.1
command:
trimseq -sequence seq.fasta -window 1 -percent 100.0 -strict -star -left -right -osformat fasta -auto
But was not able to remove those ambigous characters...
2) There are few Web-servers:
http://www.bioinformatics.org/sms2/filter_dna.html
http://www.cellbiol.com/scripts/cleaner/dna_protein_sequence_cleaner.php
They do perform the job, but they replace/remove even headers of those sequences, making the output useless...
Kindly suggest any Web-server or software/tool that could remove ambiguous bases...
Hi Rahul,
The following sed command should do the work for you.
sed '/^[^>]/s/[R|Y|W|S|M|K|H|B|V|D|N]//g' input_file > output_file
Cheers,
Eng
It would be simple to do with grep, awk, and sed.
http://matt.might.net/articles/sculpting-text/
You might want to first convert FASTA format to TABLE format so each sequence is on one line with it's name; a simple two-field table. That makes it easy to tell you script to work only on the second field (the sequence) and not the first field (the sequence name). Then convert back to FASTA format when you are done.
https://www.hiv.lanl.gov/content/sequence/FORMAT_CONVERSION/form.html
PERL also is an excellent choice for text manipulation with regular expressions etc.
Hi Rahul,
The following sed command should do the work for you.
sed '/^[^>]/s/[R|Y|W|S|M|K|H|B|V|D|N]//g' input_file > output_file
Cheers,
Eng
Some websites OR programs get a facility to remove ambiguous sequence such as BLAST encounter DUST program , the Japanese depositing facility get filtering ability to do so
Hi, U can open the fasta file as text than you can modify manually... it's depend how mutch sequences do you have.
But I need to know if there is solution to reduce the number of characters less than 50 per line?
Thank you Brian Thomas Foley
This web server works like a charm! It can replace ambiguous characters by 'N'
https://www.hiv.lanl.gov/content/sequence/FORMAT_CONVERSION/form.html
In Mothur command line you can gave commands grep and awk it works very well in my sequences.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics
- Bioinformatic Tools