Simple scripts for circularizing alignments to a linear reference
This repository contains scripts for mapping reads to the circular reference
using bwa, samtools, and a custom python script circ_fix that uses
pysam to manipulate aligned reads.
The strategy used here is to align single reads to the original reference
sequence and a modified version where some of the sequence from the start is
appended to the end of the sequence. The circ_fix script then takes the
reads that map across the end and start of the original sequence and splits
them into two segments at this point and moves the second half to the beginning
of the sequence. These new segments are added to the alignments to the original
(linear) reference, after removing any existing alignments for these reads.
These scripts use bwa but any alignment tool that produces bam files
could be used in its place. The circ_fix script works well enough on human
mtDNA sequences, but has not been tested extensively, especially in cases
where indels and clipping span the end and start of the reference. The shell
scripts included here (circ_index and circ_aln) are meant more as templates
for how circ_fix can be used rather than general purpose scripts for all
situations.
circ_index mtDNA.fa
circ_aln mtDNA.fa reads.fq reads.mtDNA