search

wyang17 / SQuIRE

Software for Quantifying Interspersed Repeat Expression
Other
49 stars 30 forks source link

Count does not consider strangeness #92

Open fgualdr opened 1 year ago

fgualdr commented 1 year ago

Hi,

I might be wrong but the counts reported by Squire seems to ignore the strandness. I checked and I am correctly specifying that the library is of type "2".. Nevertheless when comparing the counts with alternative counting methods as summarizeOverlaps the counts differ in a way that it seems to count as un stranded..

Has anyone observed the same

Thanks for the help, Francesco

Zha0rong commented 1 year ago

Hi Francesco, I am also using the SQUIRE software to do the analysis. I think the SQUIRE Count.py does pass the strandness information to the STRINGTIE: the function Stringtie() in Count.py will take in the number you entered and transform it to "", "--rf" or "--fr", depends on your input. I haven't used summarizeOverlaps before but I have tried to use HT-Seq and FeatureCounts to do the counting and they give different results from Squire-Counting on the human refGene side. I am wondering if you use featureCount or HT-Seq to do the counting do you also see different results? I am looking forward to discussing this more with you. Yours, Zhaorong