This script converts a paired-end bam file to a pairs file.
git clone https://github.com/4dn-dcic/pairix
## add pairix/util/bam2pairs to PATH.
- samtools (https://github.com/samtools/samtools)
- bgzip (this repo or https://github.com/samtools/tabix)
- pairix (this repo)
The input bam file is paired-end and it doesn't have to be sorted or indexed.
The output file contains all of the mapped reads in the input bam file. Both of the mates must be mapped. No additional filtering is performed. Every line of the bam file that meets the criterion of an 'upper triangle' (mate1 < mate2) is added nonredundantly to the output pairs file. The output file is an upper-triangular, chromosome-pair-block-sorted, pairs file, with the following columns.
readID chr1 pos1 chr2 pos2 strand1 strand2
It is bgzipped by bgzip and indexed by pairix on the chromosome pairs.
bam2pairs [-l]|[-c <chromsize_file>] <input_bam> <out_prefix>
-l: leftmost position is used (default: 5'position on the read)-c <chromsize_file>: mate ordering in the order of chromosomes in the chromsize_file is used and chromsize headers are ordered in the same way. (defautl: alphanumeric ordering of chromosomes defined in the bam header)
The first few lines of an example output file looks as below:
## pairs format v1.0
#sorted: chr1-chr2-pos1-pos2
#shape: upper triangle
#columns: readID chr1 pos1 chr2 pos2 strand1 strand2
#chromsize: chr1 249250621
#chromsize: chr2 243199373
#chromsize: chr3 198022430
...
#command: bam2pairs /d/bam/_1_out.sorted.bam /d/pairs/_1_out.sorted
SRR1658581.31870055 chr1 15398 chr1 53692634 - +
SRR1658581.24590805 chr1 19593 chr1 11784792 + +
SRR1658581.14238118 chr1 24349 chr1 67077 - -
SRR1658581.5571806 chr1 56190 chr1 56543 + -
SRR1658581.49762205 chr1 106813 chr1 252815 - +
SRR1658581.25376859 chr1 114423 chr1 143957748 - -
SRR1658581.30418686 chr1 134451 chr1 212858250 - +