Replace (optionally) blastx with Seqr

[averagehat]

See NCBI-Hackathons/seqr#19

Seqr should run a lot faster than blastx. Seqr is working, but doesn't take all the command-line options that blast does yet (and some aren't applicable, like word_size, because they're alignment parameters and Seqr doesn't compute alignments). Additionally, Seqr doesn't produce an e-value field.

Seqr should perform much faster than blastx. What commandline arguments does it need to support in order to replace blastx? See NCBI-Hackathons/seqr#22

[necrolyte2]

Do we know how it compares to diamond

[averagehat]

Diamond still computes an alignment, so Seqr should be significantly faster.
Seqr will also be able to replace blastn at some point.
But if Diamond is not "too slow" there is less incentive to use Seqr as an
alternative to blastx/diamond right now.

On Mon, Aug 17, 2015 at 10:18 AM, Tyghe Vallard [email protected]
wrote:

Do we know how it compares to diamond
https://github.com/bbuchfink/diamond

—
Reply to this email directly or view it on GitHub
#279 (comment)
.

[necrolyte2]

@averagehat I'm going to let you decide if we should keep this open or not based on your expectations of the seqr project's completion/stability