stand alone blast display header in output file (customizing -outfmt 6)

Question

[BCArg]

Dear all,

I am using standalone blast to blast a few hundred proteins against a custom database that I created (leido). I want a tabular output (displaying only the best hit) and the option -outfmt 6 is the best tested. I am using the following command to run it (pipping the output to gedit):

blastp -query subset.fasta -db leido -outfmt 6 -max_target_seqs 1  |gedit -

the output that I get (for this example, I am only submitting 3 proteins test wise (subset.fasta) follows and I know I am missing a few columns here - probably due to pasting issues in this text dialog, all columns appear on my gedit output though):

LmjF.36.6910    LdBPK_367240.1  98.88   537 6   0   1   537 1   537 0.0 1089
LmjF.36.6650    LdBPK_366960.1  96.75   553 18  0   1   553 1   553 0.0 1082
LmjF.36.5620    LdBPK_365870.1  96.77   1082    35  0   18  1099    19  1100

Is there a way to pass any flag to display the headers of the columns displayed in my output?

I tried already specifying the columns I want with (although I do not know whether or not this option should display the headers):

blastp -query subset.fasta -db leido -outfmt 6 "qseqid sseqid pident qlen length mismatch gapope evalue bitscore" -max_target_seqs 1 |gedit -

and I get the following error:

Error: Too many positional arguments (1), the offending value: qseqid sseqid pident qlen length mismatch  apope evalue bitscore.

I can easily add the columns names in R, though I would like to explore all the standalone blast functionalities

Thanks

Answer 1

[Andrzej Zielezinski]

blastp -query subset.fasta -db leido -outfmt "7 qseqid sseqid pident qlen length mismatch gapopen evalue bitscore" -max_target_seqs 1

This will give you:

# BLASTP 2.2.28+
# Query: AK364823
# Database: leido
# Fields: query id, subject id, % identity, query length, alignment length, mismatches, gap opens, evalue, bit score
# 1 hits found
AK364823    AT1G06390.2 91.23   285 285 25  0   0.0  563
# BLASTP 2.2.28+
# Query: AT1G09840.1
# Database: leido
# Fields: query id, subject id, % identity, query length, alignment length, mismatches, gap opens, evalue, bit score
# 1 hits found
AT1G09840.1 AT1G09840.6 100.00  285 285 0   0   0.0  600
# BLAST processed 2 queries

[BCArg]

I'm actually interested in the -outfmt 6 format. But yes, I have to include the 6 inside the quotes. Blastp did not complain this time, though no header was displayed

[Andrzej Zielezinski]

Output format 7 (outfmt 7) goes with the header. Format 6 is the same, but without the header.

[al-ash]

Output format 6 = tabular while output format 7 = tabular with comment lines so the difference is not only the header.

Answer 2

[shenwei356]

blastp -query subset.fasta -db leido -outfmt  "7 qseqid sseqid pident qlen length mismatch gapope evalue bitscore"

[BCArg]

I'm actually interested in the -outfmt 6 format. But yes, I have to include the 6 inside the quotes. Blastp did not complain this time, though no header was displayed

Answer 3

[al-ash]

I think there is no option to add header to outformat 6. Ugly workaround:

cat <(printf "qseqid\tsseqid\tpident\tqlen\tlength\tmismatch\tgapope\tevalue\tbitscoren") blast_output_file > blast_output_file_header