CIGAR string

[fhach]

Any plans on adding CIGAR strings to SAM output so I can load the alignments in a genome browser?

Originally posted by @yhoogstrate in #4 (comment)

[yhoogstrate]

Since the CIGAR's N blocks do not support negative numbers (for back splicing), it is in practice impossible to use one SAM segment to describe the alignment of a back spliced read entirely.

I believe that it is common practice to split a backspliced read in two and use two aligned segments of which the counter-part is soft-clipped.
Of those two segments, one (typically last part of the read) is set to 'supplementary/secondary alignment'. Both split alignments/segments then should have an 'SA' tag in place to reference each other. (http://samtools.github.io/hts-specs/SAMtags.pdf ; page 4)