extract reads from Blast output

[flacchy]

Hi...

I hope that someone can help me out with this one.

I ran blast against a db I created (specifically is from an organism I want to remove from my reads). So I ran the blast analyses and now I would like to extract all the reads and sequence that didn't have hits.

is this possible and how can I extract the reads I want from the blast output?

Thanks

F.

[okorist]

I suggest that you add the option "-outfmt 7" when running blast. This would make the result look like this:

# Fields: query id, subject id, % identity, alignment length, mismatches, gap opens, q. start, q. end, s. start, s. end, evalue, bit score
5 # 157 hits found
6 sp|Q91G65|032R_IIV6 gi|15078745|ref|NP_149495.1| 100.00 100 0 0 1 100 1 100 6e-64 198
7 sp|Q91G65|032R_IIV6 gi|291396174|ref|XP_002714758.1| 36.56 93 58 1 9 100 60 152 4e-15 74.7
8 sp|Q91G65|032R_IIV6 gi|296198107|ref|XP_002746567.1| 35.48 93 59 1 9 100 60 152 5e-15 74.3
9 sp|Q91G65|032R_IIV6 gi|116004445|ref|NP_001070581.1| 36.56 93 58 1 9 100 60 152 8e-15 73.9
10 sp|Q91G65|032R_IIV6 gi|126309825|ref|XP_001370260.1| 36.56 93 58 1 9 100 60 152 1e-14 73.6

Then it would be easier to extract the reads from the output.

[flacchy]

I used a script to obtain the same output ... the issue is extract them...

[rhinoceros]

I used a script to obtain the same output ... the issue is extract them...

blastdbcmd, read the manual..

[flacchy]

I've read the manual but I found more useful use the script ncbi_parse.pl because I needed the description of the organism before and the tab output don't gave that infromation (I saw also in some blog other people having the same issue) I was only wondering if there was a way to extract the reads and sequence of the no hits from the output...

[rhinoceros]

I've read the manual but I found more useful use the script ncbi_parse.pl because I needed the description of the organism before and the tab output don't gave that infromation (I saw also in some blog other people having the same issue) I was only wondering if there was a way to extract the reads and sequence of the no hits from the output...

With blast 2.2.28+ you can have organism name in the output given you have setup taxdb properly and use e.g. -outfmt '6 std sscinames'..

[GenoMax]

flacchy wants to extract reads that did not have a blast hit in results. Not sure if blastdbcmd allows that.

[rhinoceros]

flacchy wants to extract reads that did not have a blast hit in results. Not sure if blastdbcmd allows that.

Ah yeah, my bad. So for tsv/csv output:

cut -f 1 (optional -d ',' for csv) yourResultFile | sort -u > seqIdsYouWantToRemove.txt. Then ideas for that from for example here..

[yzzhang]

once you get the reads list you want to remove, you can use comm command to get left reads names wanted, and then extract them very easily. fastqselect.tcl from mira software works very well