Running Blast For A List Of Pairs

Question

[Freddie]

Hi,

I would like to ask if it is possible that I run BLAST with a list of pairs of query and reference sequence IDs?

I tried bl2seq but unfortunately it does not produce the alignment of the whole length. So, for example, if a pair of sequences can produce 4 alignments, then bl2seq will output 4 results for each alignment, each with its own percentID value, while blastall extends and joins all the alignments together, and so it produces only one value of percentID of the whole. I would like to be able to run like this blastall but for all pairs that I have selected whose ID I have put in the list. Is this possible?

Thanks heaps.

[Freddie]

by the way, this is for protein sequence.

Answer 1

[Joseph Hughes]

You could do a blastall of the IDs you are interested in against database of the reference you want to use. Create a fasta file with you sequences to compare to, e.g. reference.fa. Format the database:

makeblastdb -in reference.fa

Then blast your set of interest against this database

blastp –db input.fa –query input.fa –out results.out

The output will provide you with the percent identity for the length of your query.

Answer 2

[Hamish]

In NCBI BLAST+ the pairwise alignment capability that was provided in 'bl2seq' has been merged into the main programs (e.g. blastp, blastn, blastx, etc.). For details of how to use the NCBI BLAST+ (and legacy NCBI BLAST) programs to perform pairwise alignments see:

• Standalone Blast 2 Short Sequences
• Correct Method To Blast All-Vs-All With Ncbiblast & How To Speed It Up?
• Compare Two Protein Sequences Using Local Blast

And many of the other BLAST related questions on Biostar: blast

Depending on your requirements the pairwise local alignment provided by NCBI BLAST and NCBI BLAST+ might not be the most appropriate method. So you may want to look at the following:

• Local/local alignment: alignment of local regions with high similarity
  • EMBOSS programs: water, matcher and supermatcher
  • FASTA suite programs: fasta, fastx, fasty, lalign and ssearch
• Global/global alignment: alignment from end-to-end of both sequences
  • EMBOSS programs: needle, needleall and stretcher
  • FASTA suite program: ggsearch
• Global/local alignment: local region in one sequence against the whole length of the other
  • EMBOSS program: est2genome
  • FASTA suite program: glsearch

For more options see the List of sequence alignment software page in Wikipedia.