HT Seq Count stranded options

[qwrissie]

I am very new to bioinformatics, so I would be really grateful for some help!

I have been using HTSeq Count v0.5.3 and I am bit confused about how the stranded options work. I have used both -s no and -s yes and for a few genes the number in the stranded set are higher than in the non-stranded set (and I haven't changed anything else). Looking in the genome browser, I think this is because in -s yes mode those reads for which the mate maps to a different gene or maps ambiguously are counted, whereas -s no ignores these. My data is stranded, so I want to use the stranded mode, but if that means that I am including reads where the mates map improperly, am I better off excluding these from the data before I run HTSeq Count?

Also, HTSeq Count does not appear to be counting singletons. Is there a way to change that?

[Simon Anders]

If you have features on different strands that over close by or even overlapping then a read that overlaps with both features will be discarded with '-s no', because is cannot be assigned to only one of the two features, but with '-s yes' it will be counted for the feature on the same strand as the read. Does this explain your observation?

And what do you mean by "singleton"?

[qwrissie]

Yes it does! Thank you very much for your swift reply, you have saved me a lot of time. By singleton, I mean reads for which the mate has not mapped.