An option to output SEQ field for secondary alignment

[fenderglass]

Hi!

I have been using minimap2 as a part of the Flye pipeline. I am using secondary alignments during consensus/polishing to account for possible duplications in disjointigs and improve the base quality of long repeats. Storing SEQ field for secondary alignments in SAM/BAM files makes the alignment file parsing much easier, since all separate alignments could be processed independently.

An existing -Y option solves the problem, but it also forces supplementary alignments to use soft clipping, which in some datasets dramatically increases alignment size. I therefore added a new option -secondary-seq that enables output of SEQ for secondary alignments, and uses hard clipping for both supplementary and secondary alignments.

The option was extensively tested as a part of the Flye pipeline for almost a year. I however have not tested outside of genome assembly setting.

Best, Mikhail

[tillea]

It would be really great if this patch could be applied to enable building flye smoothly.

[fenderglass]

Hi all,

Will be happy to make this pull up-to-date with the master, once you confirm that you are interested in adopting. The conflict is due to new command line options, so it is an easy fix.

[olechnwin]

@fenderglass, I don't know how the pull works. But, I just posted the same request here. It would be great if sequences for secondary mapping can be turned on. Thanks!

[cmdcolin]

just saw this PR but I made a program that tries to help add SEQ to secondary alignments as a post-processing step though this PR would be interesting too! https://github.com/cmdcolin/secondary_rewriter