Nanopore · Plotting · Python

Insertion-deletion ratio in Oxford Nanopore PromethION data

It’s been a while since I wrote another blog post, so here is something short. Today I had to figure out the inversion deletion ratio in Oxford Nanopore Sequencing data from PromethION human whole genome sequencing.

I want to compare this ratio for reads which are qualified by albacore as “pass” (average base call quality score above 7) and “fail”(average base call quality score below 7). For speeding things up, I only used chromosome 22 for this.

samtools view anon_pass.bam chr22 -o anon_pass_chr22.bam
samtools view anon_fail.bam chr22 -o anon_fail_chr22.bam

To get the insertion-deletion ratio I will use the CIGAR string from the aligned reads. The Python script can be found below. The code uses pysam and iterates over the bam file, using the cigartuples attribute. In this attribute, insertions get a “1” and deletions a “2”as the cigar operation code (op). The other part of the tuple is the length of the operation. So a tuple of (1,25) corresponds to an insertion of 25 nucleotides relative to the reference genome.

For deletions, the minimum is 1 to avoid divide by zero errors. The boolean expression on line 18 will evaluate to 1 if the first part evaluates to 0.

From these ratios I create a simple histogram using matplotlib, using bins of 0.1 with a maximum of 5. The resulting plots are shown below. So the ratios are:

pass: 1.002
fail: 1.027
combined: 1.012

Indeed showing a difference between failed and passed reads. That’s all for today!

Insertion Deletion ratio for “fail” reads
Insertion Deletion ratio for “pass” reads

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s