How to filter BLAST results by query coverage? How to analyze large BLAST outputs?

Question

[braun_tube]

Hello,

I am currently using blastn. I can use the perc_identity parameter to set my percent identity cutoff to 80%. However, I also want to set an alignment length / query coverage cutoff of 80%. Is this possible?

I am also using tblasn which does not even have a perc_identity parameter.

In the absence of these filters is there some way to easily analyze a large blast output? For example I did a tblastn on 100 isolates looking for hits of a specific gene. Since tblastn does not have a percent identity filter or alignment length filter I have lots of hits above and below my threshold of 80%. Is there a way for me to quickly check how many hits I have above my threshold without manually checking the values for each hit in my blast results?

[mthm]

to get query coverage with blastn use -qcov_hsp_perc

Answer 1

[young_bioinformatician]

After getting the results, you can use awk command on linux to filter according to any column.

Example:

awk '$3>80 {print}' file > filtered_file

for two conditions:

awk '$3>80 && $11 == 0 {print}' file > filtered.file

[braun_tube]

Thank you, this worked perfectly.

Answer 2

[brian.fristensky]

Have a look at the BIRCH system, which gives you graphical interfaces for running BLAST searches, and for working with the output in spreadsheet form. The Biolegato spreadsheet lets you quickly go through hundreds of hits, and directly retrieve the sequences you choose into sequence sequence tools for downstream analysis.

I just posted links to two new videos yesterday:

VIDEO: BLAST Searches Through a Data Science Lens

VIDEO: Installing BLAST Databases on Your Own Computer