Blastn output retrieve sequence

[rkizen]

Hello, is there an easy way to retrieve the sequences from the hits in the blastn output? I just need the sequences without the gap insertions "-". Thank you.

[GenoMax]

Look at the "blastdbcmd" program examples on this page. http://www.ncbi.nlm.nih.gov/books/NB...BLAST_search_r .. if you are looking to retrieve the subject sequences.

[rkizen]

Thank you GenoMax. I tried using blastdbcmd but for some reason it doesn't seem to be working as described on that page.

I have downloaded the Gallus 4 genome from NCBI and ran makeblastdb using blast+ 2.2.25. When I run:
blastdbcmd -db gga_ref_Gallus_gallus-4.0_all.db -entry all | grep ">" | head
I get:
>gnl|BL_ORD_ID|0 gi|358485511|ref|NC_006088.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 1, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|1 gi|358485510|ref|NC_006089.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 2, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|2 gi|358485509|ref|NC_006090.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 3, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|3 gi|358485508|ref|NC_006091.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 4, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|4 gi|358485507|ref|NC_006092.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 5, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|5 gi|358485506|ref|NC_006093.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 6, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|6 gi|358485505|ref|NC_006094.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 7, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|7 gi|358485504|ref|NC_006095.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 8, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|8 gi|358485503|ref|NC_006096.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 9, Gallus_gallus-4.0, whole genome shotgun sequence
>gnl|BL_ORD_ID|9 gi|358485502|ref|NC_006097.3| Gallus gallus breed Red Jungle fowl, inbred line UCD001 chromosome 10, Gallus_gallus-4.0, whole genome shotgun sequence

But when I try running:
blastdbcmd -db gga_ref_Gallus_gallus-4.0_all.db -entry 358485511

I get:
Error: 358485511: OID not found
BLAST query/options error: Entry not found in BLAST database

The entry seems to be there so why can't blastdbcmd recognize it when I try to query it?

[rhinoceros]

Did you run makeblastdb with '-parse_deflines' ?

[rkizen]

Ok I just ran makeblastdb with the -parse_seqids option and it works now. What is the difference between parse_seqids and parse_deflines?

[rhinoceros]

Ok I just ran makeblastdb with the -parse_seqids option and it works now. What is the difference between parse_seqids and parse_deflines?

Sorry, -parse_seqids is the flag I was thinking..

[maubp]

See http://blastedbio.blogspot.jp/2012/1...cbi-blast.html for my thoughts on the difference between the NCBI style IDs and free-format user provided identifiers (and the problems using them with blastdbcmd).