Hi. I'm trying to get the read count by cufflinks, as I need the read count of each transcript so I can't use HTseq. I remember cufflinks used to report raw read counts. Can someone suggest me how can I get raw read counts by cufflinks?
They aren't raw counts. Cufflinks computes expression in a probabilistic way to infer the read/transcript association.
If you can get reads out of cufflinks, they surely won't be "raw reads".
A tool like HTSlib will help you count things.
HTseq is a good tool for read count, cufflink only does differential expression analysis.
Cufflinks is used for transcript assembly from aligned files. If you want to check whether distribution of reads over a co-ordinate support the gene model.
featureCount has served best for me to get read counts.
hth
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Although I'm not a big fan of Cufflinks, it can do a bit more. It computes based on some coverage-assumptions the read distribution per gene locus and assigns FPKM-values to the genes' transcripts.
I would not recommend trying to recompute the read counts from the FPKM values, since Cufflinks does some length normalisations of the transcripts as well as assigning multi-mapping reads.
However, I also recommend using HTSeq-count for getting the read-count.