Select only sequences that align with X% of a query during blastp search

Question

[dago]

I have a Multifasta file that I use as query in a blastp search against nr.

I would like to set a lower threshold indicating that the blastp hits need to be aligned to at least x% (e.g. 80%) on the query sequences. I could not find any option in blast that allows that.

Am I missing something?

[iraun]

I'm not sure if you can set the threshold before the analysis, but, you can filter the output according to your needs. You can customize the output using -outfmt parameter, and select those hits having the percentage of the hit over the length of the query (length/qlen) >= 80%.

[dago]

thanks for the comment. Basically I have to customize the -m 6? I cannot see the option though. I guess it would be this one "qcovs means Query Coverage Per Subject"

[Juke34]

Yes it is

BLAST definition and difference between 'qcovs' and 'qcovhsp'

[dago]

OK, I think I am dumb. I try to set this option in the -m flag, but only the defaults values are displayed.

blastall -p blastp -d ~/Documents/Blast_DBs/NR  -i query.fasta -e 0.0000000001 -m "8 qseqid sseqid evalue qcovs"  -a 20 -o NR_15-5-15

[Juke34]

I don't really understand... you talk about the qcovs option and you are using blastall. Something is wrong.

If you are using the old BLAST (blastall) the option is -m 8. You cannot define what you want in the different columns. You will have : 1) query / 2) subject / 3) % id / 4) alignment length / 5) mismatches / 6) gap openings / 7) query start / 8) query end / 9) subject start / 10) subject end / 11) e-value / 12) bit score

But it's really encouraged to use Blast+ instead, and the command will be blastp -db database -query Input.fa -outfmt "6 qseqid sseqid evalue qcovs".

I think you are mixing both tool blast and blast+. In your case you must use the blast+ to retrieve the qcovs value.

[dago]

Got it! I am quite confused with the differences between Blastall and Blast+. what does it change?

[Juke34]

Lot of things. Mainly, speed improvement, they added new parameters (by the way they changed the names of old parameters). You can also call directly the program you need e.g. (blastp parameters. ; blastn parameters) :

blastall parameters: http://www.compbio.ox.ac.uk/analysis_tools/BLAST/BLAST_blastall.shtml

blast+ parameters: http://www.biomedcentral.com/content/supplementary/1471-2105-10-421-s1.pdf

To know more about what has changed read:

• http://www.biomedcentral.com/1471-2105/10/421
• http://wiki.christophchamp.com/index.php/BLAST%2B

[dago]

Thanks a lot!