Handling of supplementary alignments

[porchard]

Hi,

Thanks for the tool! I recently ran this on a large cohort and noticed some oddities in the output (e.g., end 1 mapping rates > 1 for a subset of the libraries). The issue seems to be attributable to a large number of supplementary alignments. As far as I can tell RNA-SeQC doesn't check the 0x800 flag denoting supplementary alignments and this throws off some of the numbers -- for example, inflating the 'unique mapping, vendor QC passed reads'.

Given the set of metrics that RNA-SeQC calculates, I would have expected it to exclude supplementary alignments from many of the metrics as it does for secondary alignments. Is there any particular reason for ignoring the supplementary alignment flag?

Happy to put together a pull request to treat supplementary alignments in a similar manner as secondary alignments, if you determine that to be desirable.

[agraubert]

Thank you for bringing this to our attention! We apologize for not catching this issue earlier. We're currently working on adding in an appropriate filter for supplementary reads and will keep you updated

[porchard]

That sounds great -- I'll look forward to the updated version, then!

[agraubert]

Just wanted to give you an update. We've updated the way RNA-SeQC counts chimeric reads and how it handles supplementary alignments. We're testing the new changes and expect to finalize the new changes soon

[porchard]

Thanks! I'll give it a go once the changes are pushed.