Some Alignments Have Impossible Flag Values

[DarioS]

I mapped whole genome sequencing data using the -a option. Some alignments have a Flag total of 377 and 441. According to Explain SAM Flags internet tool, 377 includes mate unmapped, read reverse strand, mate reverse strand. How could the mate be unmapped, but also on the reverse strand? Similarly, 441 has the same issue of being contradictory. I came across this by chance because I was trying to decide what value to provide to samtools view -f to explore all of the alignments added by the -a option of bwa mem. It could be more explicitly documented, similar to #206 request.

[quokkamole]

I've encountered the same problem.

I have paired reads where one has a SAM flag total of 117 and the other has a value of 185. In both cases, these values indicate that the unmapped read is associated with the reverse strand.

The alignment were bwa mem PE alignments with -M flag.

Edit:

This situation also seems to be described here:

https://www.cureffi.org/2012/12/19/forward-and-reverse-reads-in-paired-end-sequencing/

and here:

https://www.biostars.org/p/94213/