deseq dataset from star

Question

[Pals]

Hi,

I have multiple count tables generated by star (using --quantMode). Is it possible to import these tables into deseq2 using DESeqDataSetFromHTSeqCount function or I need to create a matrix table and import using DESeqDataSetFromMatrix function? the tables look like this:

N_unmapped  2399497 2399497 2399497
N_multimapping  5671935 5671935 5671935
N_noFeature 18272205    33941344    34175527
N_ambiguous 786927  66964   66285
ENSOARG00000017577  296 141 155
ENSOARG00000020430  7   5   2
ENSOARG00000017587  8   4   4
ENSOARG00000025485  3   1   2

Thanks, Kisun

[WouterDeCoster]

You probably have to remove 2 columns and those upper N_ lines to use DESeqDataSetFromHTSeqCount

Answer 1

[bruce.moran]

I usually use:

dds<-DESeqDataSetFromMatrix(counts,colData=conditions,design =~ treatment + tissue)

on a matrix I have made in featureCounts. 'conditions' is a matrix where each row is a sample ID and corresponds to a column in 'counts' (I think these have to be ordered the same), and contains multiple variables that can be used in the design, e.g. here there are two columns in the 'conditions' matrix, 'treatment' and 'tissue'. After this I just run

ddseq<-DESeq(dds)

Any questions feel free to ask.