How can one get raw read counts from RPKM values

[gen2prot]

Hi All,

I am new to the transcriptome sequencing field and I am learning to use Bowtie, Tophat and Cufflinks. I find that Cufflinks spits out RPKM and FPKM values which though very informative does not help my purpose since I am interested in the number of raw reads per gene. That is to say without normalization. Is there a way of figuring this out?

Thanks
Abhijit

[mgogol]

One way I've done it is using bedtools coverageBed with the bam file generated from the sam file from the tophat alignment and a bed file describing the genes.

I am curious to hear other people's techniques, because I'm sure there are more efficient ways to go about this.

[lexa]

hi,

I`m also using bedtools but does anybody has a better solution?

[edge]

Hi,

Do you mind to share the command that you type in bedtools in order to extract the read count per transcripts?
I'm facing the same problem as well.
Thanks.

[edge]

Hi,

Do you mind to share the command that you type in bedtools in order to extract the read count per transcripts?
I'm facing the same problem as well.
Thanks.

[gen2prot]

Getting raw read counts

I ended up using Simon Ander's program for computing read counts per gene. However you will have to provide the annotation of the genome in GTF format in order to view that. I find that this is quite accurate. Cheers...

[biofreak]

I tried Htseq-count software.
python -m HTSeq.scripts.count accepted_hits.sam hg19RefGene.gtf

it works. But I want to do this in R. Maybe functions in the GenomicRanges packages can help? e.g. countOverlaps function?