Error when retrieving fasta file

sequenceserver
#1

I got this error when clicking on the fasta for a search result:

Thanks in advance!

#2

This error happens when there’s stop codon (*) in the protein sequence database.

Priyam

#3

Well that was an easy fix. Thank you!

#4

Hello Daniel and Priyam, I have just reinstalled sequenceserver 1.0.11 and am getting this same error when attempting to retrieve fasta file hits. You stated why this may happen, but how do I fix the problem (specifically) . . . no bioinformatic skills in my toolkit!

Thanks, Michael

Something went wonky

Looks like you have encountered a bug in SequenceServer. Could you please report this incident to your admin or to SequenceServer Google Group (if you are the admin), quoting the error message below?

#5

It looks like the actual bug is in blastdbcmd from NCBI:

$ blastdbcmd -outfmt ‘%g %i %a %t %s’ -db ‘prot.fa’ -entry ‘F1’

→ …CSCASIFÿÿQMFPSQFRDGPSITWSÿSNRAKSASVÿKASSSÿEGI…

Where every Stopcodon ‘*’ in the prot.fa is replaced by ‘ÿ’ by blastdbcmd (0xFF).
So replacing it in the output, fixes the issue.

So in sequence.rb (in my case in /var/lib/gems/2.1.0/gems/sequenceserver-1.0.11/lib/sequenceserver)
in Line 189-191 add the following:

command = “blastdbcmd -outfmt ‘%g %i %a %t %s’”
" -db ‘#{database_names.join(’ ‘)}’"
" -entry ‘#{sequence_ids.join(’,’)}’ | tr ‘\377’ ‘\052’"

This will translate 0xFF back to 0x2A and everything is fine.
This is probably better done later on and directly in ruby instead of piping through tr but i don’t know ruby well enough to do it the proper way and this works for me :wink: