This is a test version of Biostars. For the public version, visit https://www.biostars.org.
megablast returns hits not found by blastn?

Hi all, It is my understanding that blastn is more sensitive (and thus finds more hits) than megablast (see here and here). However, when I compared my results from running blastn vs. megablast with the same exact inputs and parameters, I got 0 hits for blastn and thousands of matches for megablast (with e-value less than 0.001).

This seems really strange to me and I was wondering if anyone can think of a specific case in which megablast picks up hits that blastn can't pick? These are the commands I ran: 

 blastn -query unique_31mers.fa  -db Homo_sapiens.GRCh37.dna.primary_assembly  -outfmt "6 delim=\t" -task 'megablast'  > GRCH37_blastn_31mers_mega.out

blastn -query unique_31mers.fa  -db Homo_sapiens.GRCh37.dna.primary_assembly -task 'blastn'  -outfmt "6 delim=\t"  > GRCH37_blastn_31mers.out

For context, the 31mers come from viral sequences.

Thanks!

Best, Elaine

blastn megablast e-value query blast+
• 1,870 views • link
written by 7ehuang • 7ehuang

1 answer

It is not clear what result you got here because you are saying I got 0 hits for blastn and thousands of matches for blastn. I would expect that you are getting 0 matches with megablast because of its larger default word size.

Either way, you may want to try -task blastn-short as this is optimized for sequences shorter than 50 nucleotides.

• 0 views • link
written by Mensur Dlakic Mensur Dlakic

Sorry about that, I mistyped. I meant to say I got 0 hits for blastn and thousands of matches for megablast (edited the original post). This is why I'm confused, because I also expected getting more matches with blastn than megablast. When I tried -task 'blastn-short', I also got 0 matches.

• 0 views • link
written by 7ehuang • 7ehuang

Log in to answer this question.

More posts like this