What Makes A Valid Gi List?

Question

[Andrewktmeikle]

Hello all, I cannot figure out what makes a "correct" GI list!

Atm I have a list that I produced this using the query below and a awk line to just retrieve the second column values

blastn -word_size 11 -reward 2 -penalty -3 -gapopen 5 gapextend 2 -query hiv_reference.fa -db test -outfmt 6 >results.txt

awk "print $2" results.txt > giResults.txt

G3ME69S02DME82 G3ME69S02EPWTE G3ME69S02DE0PT G3ME69S02C1ABI G3ME69S02EQ3LK G3ME69S02ELVHG G3ME69S02CY21D G3ME69S02D9A53

So that is what I get from running those two lines, and from what I understand those are GI values, so why can I not do

blastdb_aliastool -gilist giResults.txt -dbtype nucl -db test -out test_subset

I get an error that its not a valid GI list.

Any ideas?

Any input is much appreciated

[boczniak767]

Maybe this topic will help you (especially my comments ;) ).

http://biostar.stackexchange.com/questions/1196/extracting-sequence-from-a-3gb-fasta-file

Answer 1

[Neilfws]

from what I understand those are GI values

Those are not GI values. A GI is an identifier assigned to a sequence record by the NCBI. It's simply a series of digits such as 12345 (in that case, a protein sequence record).

In a BLAST report, this would look like:

gi|12345|

What you have is some other kind of sequence identifier which is not a GI.

[Andrewktmeikle]

Excellent, thank you very much I'll look into that I report back with how I get on! :D

[Andrewktmeikle]

Yeah i think they are all sequence ID's, so is there a special type of query you have to do in order to get the GI's back?

[Neilfws]

Well, where did those sequence IDs originate? When I search for them in Google, this question is the only hit. So I doubt that they are linked to GIs in any publicly-available resource.

Answer 2

[Iván]

I haven't used a gilist in BLAST, but maybe this link will be of help, although it's not for BLAST+ exactly. Check out specially the "Alias file structure" bit. Hope it helps!