calculate RPKM from read count and gff by galaxy

Question

[maleknias]

Hi ! I have a table (58037×117) contain scaled read count for 117 sample (case and control) and gff and gtf file. Considering of RPKM calculation: RPKM=(row number of reads/exon length) ×(1000 000 / number of mapped reads in the sample) how can I get a table (58037×117) contain RPKM?? Is it solution based of use galaxy?

Answer 1

[Matteo Schiavinato]

In the original publication by Ali Mortazavi et al. they provide the means to do so: http://www.nature.com/nmeth/journal/v5/n7/abs/nmeth.1226.html?foxtrotcallback=true

If you are searching for a more "divulgative" article on the topic, this is the most linked one: https://haroldpimentel.wordpress.com/2014/05/08/what-the-fpkm-a-review-rna-seq-expression-units/

Finally, a remark: I would suggest you to switch to TPM. Mortazavi himself suggested not to use RPKM anymore.

[maleknias]

Thanks but it's not useful for me. I want to save time and because sample size is big i want use table of read-count. so do it again from the beginning is not cost effective!!

[Matteo Schiavinato]

You don't need to do everything from the beginning, it just takes some command line used differently because the formula is different.

Answer 2

[EagleEye]

1) Enter "#" for your matrix header

2) Add one extra column to your matrix in the end (exon/gene length)

3) Use the following script according to the instructions,

https://github.com/santhilalsubhash/rpkm_rnaseq_count

[maleknias]

thanks a lot. It seems cool. could you guide me "how I can make .count file" ?? Is this make by "Count.seqs" and "mothur_toolsuit" ?

[EagleEye]

count file can be any matrix of raw count (Gene id in the first column+followed by counts+followed by Gene/exon length) created by read quantification programs like 'featureCounts' and 'HTSeq'. For getting sum of exon length gene-wise 'featureCounts' is well suited program.