Reporting of multiple alignments using hisat2

Question

[markus.glass]

Hi,

I would like try HISAT2 for aligning RNA-Seq data but I am uncertain about using the -k parameter for reporting N multiple alignments. The manual says a primary alignment is an alignment whose alignment score is equal or higher than any other alignments. But it also says HISAT2 does not guarantee that the N alignments reported are the best possible in terms of alignment score. Does this mean i) the N reported alignments may not be all alignments with the best alignment score or does it mean ii) there may be alignments with a better score than of those N reported?

i) would be clear, but in case of ii) I don't understand how this is possible if a primary alignment has an alignment score equal or higher than any other alignments.

Thanks in advance!

Answer 1

[predeus]

This is actually one of the most confusing things about alignment in general.

What this description means (I think) is that if you have -k set to 10, and aligner will find 10 locations where the read aligns with a certain mapping score, it will report them and stop looking further.

What I do not get is that, according to the description, there should only be primary alignments in the resulting SAM. That's certainly not the case: quite a few of them have "secondary" bit-wise flag (256) set.