Entering edit mode
8.5 years ago
tonja.r
▴
600
I used featureCounts of Rsubread and found something that I cannot explain:
I wanted to count the features in this region chr1:3044014-3044814, strandSpecific parameter was set to 0.
Somehow I do not understand how the read ...4974:39966 got into the results. Neither its end, nor its starts lie in the specified region.
If I set the strandSpecific parameter to 1. This read does not appear in the results.
HWI-ST227:372:C3UVCACXX:3:1113:4974:39966 16 chr1 3043884 255 50M * 0 0 CTGAGAGCTTGGCGCGCTGTCTCTGCAGCATAGGGGCACTAACTATGCCCJJJJJJIJJJJJJJJJJJJJJIJJJIJJJJJJJJJJJHHHHHFFFFFCCC XA:i:0 MD:Z:50 NM:i:0
HWI-ST227:372:C3UVCACXX:3:1212:16462:69280 0 chr1 3044215 255 50M * 0 0 TCCAAGATTGGTGCCGAGCCCTATTGGCCCAAGTTAGAAGTAGACATCTGCCCFFFFFHHHHHJJJJJIJJJJJIGGJJJJJJHHIIJHIBFHIJJJIIG XA:i:0 MD:Z:50 NM:i:0
That shouldn't happen, perhaps you found a bug. How are you checking that featureCounts is including that alignment in the count?
by
reportReads=TRUE. It produces the file with each read if it is assigned or not. And then I filter those which are assigned.Can you make a very small (perhaps just the header and this alignment) SAM or BAM file that reproduces this issue? It'd be good to get a bug report filed so this can get confirmed and fixed.
Yep, where should I send the file and featureCounts command?
There's a google group for subRead here. If you post things there then Wei Shi will see them and can hopefully confirm and fix things. Please do post back once you hear back from Wei Shi.