Makeblastdb And Blastx

Question

[Ken]

Hi, I tested the makeblastdb and blastx but doesn't seem working for me. Can you please check if I did anything wrong?

I used 2 fasta sequences (SETREF : testDB.fa) as a reference and a file with partial sequence of one of the sequences in SETREF as a query (SET_Q : testQ.fa). But the result shows No Hit.

This is the SET_REF:

>Glyma0021s00410.1|PACid:16242639
ATGGCTCGTAAGAAGCAAA[...]
>Glyma0021s00460.1|PACid:16242642
ATGGAACAAAATCTGTACA[...]

And this is the SET_Q:

>seq1
ATGGCTCGTAAGAA[...]

seq1 in SETQ is just the first line in the first sequence in SETREF.

I applied makeblastdb to SET_REF:

makeblastdb -in testDB.fa -title "test DB"

Then applied blastx:

blastx -db testDB.fa -out testQ.blastx -query testQ.fa

But then there is No Hit for the query sequence, did I missed out something? Why the query sequence is an exactly same sub-sequence as the reference but returned no alignment? Thanks a lot in advanced.

Answer 1

[ALchEmiXt]

For BLASTx you need a PROTEIN database. To this DB you can only do a BLASTn. Or do I misunderstand?

An alternative for using the DB as nt would be to use tblastx. Translated blast query against a translated DB.

[Michael Schubert]

Nothing to add here. Except, maybe: http://blast.ncbi.nlm.nih.gov/Blast.cgi?CMD=Web&PAGE_TYPE=BlastDocs&DOC_TYPE=ProgSelectionGuide#blastx

[Ken]

Thanks ALchEmiXt, I got it now. I was confussed with the makeblastdb option -dbtype. It can take 'nucl' (nucleotide) or 'prot' (protein)[default] as arguments. I misunderstand that it will translate the input fasta sequence and make the output as a protein database. Then in this case, I think I need to use '-dbtype nucl' as an option for the makeblastdb, don't I?

[ALchEmiXt]

Yes you can only FORMAT a BLAST DB of iets own content. The algorithms can ANALYSE 6 frame translated versions for you. The blast+ suite of ncbi also lets you do this directly on input FASTA query AND FASTA DB sequences.

[Ken]

Thanks heaps!!!