Parsing BLAST output. Output of a local BLAST search “less” program Full path to the BLAST output file.

Parsing BLAST output

Output of a local BLAST search

“less” program

Full path to the BLAST output file

BLAST program used for the search

Reference

Information of the query sequence

One-line summary of the search results

Detailed information for the first 2 hsps of the first hit:Accession number, description, organism, score, E value, identities, positives, and alignment

Information of the database

Sample BLAST output (continued)

Hsp information from the first hit

Press “q” to quit the “less” viewing mode

The size of the BLAST output is limited only by the free disk space you have in your computer. It’s virtually impossible to open a large text file. Let alone going through the file line by line.

The purpose of parsing BLAST output is to extract user-defined information from the BLAST output file for clear visualization and summarization.

Search result parsing

The Bio::SearchIO system was designed for parsing sequence database searches (BLAST, sim4, waba, FASTA, HMMER, exonerate, etc.)

One-line summary of the search results

Load Bio::SearchIO moduleUsage informationIt will appear if the program is invoked without arguments

Define the class

Print out the header information

Process each result

Process each hit

Process each HSP

Control for the number of hits to be extracted

Indicator showing the work is done

Change directory (cd) to where the perl script and the BLAST output file are stored

Confirm that the perl script and the BLAST output are in place

Oops… an error message

It’s due to Windows and Unix compatibility.

Find the file in Windows system and open it with Notepad++

Select “convert to UNIX format” in the “Format” drop-down menu

After the conversion, save the file and exit Notepad++

Another error message

This is because the perl interpreter has been installed in another location (/usr/bin/) while the script is looking for the perl interpreter in /usr/local/bin

Solution:Create a symbolic link of /usr/bin/perl in /usr/local/bin

Command:ln<space>-s<space>/usr/bin/perl<space>/usr/local/bin/perl

Now it’s working !

Congratulation! You’ve just parsed a BLAST output!

This is the file you’ve just generated.

Let’s see how the file looks like, using “less”.

Here is how it looks like.

The parsed output is tab-delimited and can be imported into Excel for better visualization.

Locate the file in Windows system

Accession numbers of the top 3 hits

Descriptions of the top 3 hits

E values of the top 3 hits

Information of each HSP of the top 3 hits

Header row

Query sequence