HTSeq to DESeq

[muthu545]

Hi all,

HTSeq's count file output contains # of reads for
no_feature
ambiguous
too_low_aQual
not_aligned
alignment_not_unique

I used to delete them from my gene matrix before inputting that
to DESeq's newCountDataSet.

Is it okay to do it this way or have these 5 feature_id's in the data
matrix when performing DE analysis using DESeq

[Simon Anders]

You are correct, you should delete them.

[muthu545]

Thank you, Simon.

[tschauer]

Hi Simon,

The DESeq2 vignette (HTSeg input) confused me about this. The "_lowaqual" is still in there. See

ddsHTSeq
class: DESeqDataSet
dim: 70467 7
exptData(0):
assays(1): counts
rownames(70467): FBgn0000003:001 FBgn0000008:001 ... _lowaqual
_notaligned
rowData metadata column names(0):
colnames(7): treated1fb.txt treated2fb.txt ... untreated3fb.txt
untreated4fb.txt
colData names(1): condition