Hi all,

I was wondering if I'm missing something obvious: samtools can filter your BAM file based on many criteria (such as flags, tags, qlen etc) - but what is the correct way to get rid of the chimeric mappings (at least the type where R1/R2 map to different chromosomes?). I can come up with a one-liner (decode to text/check 7th column/encode back to binary), but I think this should be in samtools - am I wrong?

I am working with hisat2 BAMS, and no flags are set from what I can tell. This is one of the mappings:

SRR12813830.114509898   65      NZ_JAHRBL010000049.1    1405    1       46M     NZ_WJPN01000041.1       1609    0       TCCTGGTGGTGCCCTTCCGTCAATTCCTTTAAGTTTCAGCTTTGCA  FFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFFF  AS:i:0  ZS:i:0  XN:i:0  XM:i:0      XO:i:0  XG:i:0  NM:i:0  MD:Z:46 YT:Z:UP NH:i:18

The "samtools" way of specifying primary alignments is to remove supplementary and secondary alignments:

samtools view -F 265 -F 2048

but as for your specific question, if I recall correctly, bwa for example has a special behavior where when a template is matching to different chromosomes it will have all POS field set but the flag will contain the unmapped flag as set. It is the weird, unmapped read, that has a position...

This is bwa specific behavior I believe.

Many aligners do not mark the chimeric and supplementary reads, so in those cases you are out of luck from - from flags alone you can't filter these, and you would need to write some processing code yourself.

Or perhaps one of the tools in the BAM file cottage industry has a prebuilt feature that suits you, try:

• samtools, picard, sambamba, samblaster, samblast, samsift, bamtools, biobambam