One of the most useful things about BINSEQ files is that cheap per-sequence tasks can be trivially parallelized. This is great, because many problems in bioinformatics belong to a class of problems that are embarrassingly parallel.

One such task is searching for specific subsequences within each sequence. This is equivalent to performing a grep command on each sequence.

This is such a common operation that bqtools provides a built-in command for it: bqtools grep.

Let’s look for the subsequence ACGTACGT in our records.

Looks like we got some hits! You’ll see that bqtools will highlight the matches for you in the sequence.

Let’s actually just count the number of records we got with this search using the -C/--count flag.

We can also add additional patterns to search for in our records. By default the logic for accepting a read with multiple patterns follows an AND logic, which means that a read must contain all patterns to be accepted.

Let’s find all the records that contain both ACGTACGT and TATA and output as a fastq using the -f/--format option:

bqtools grep is a fully Regex compliant tool, which means you can use regular expressions to search for sequences.

Let’s take a look at a more complicated query - perhaps a sequence that begins with ACGT but then has either TA or GC, and finally ends with CCGG.