Entering edit mode
8.2 years ago
aswartz85
▴
20
I'm just trying to make a plot for particular genes (~200) showing the correlation, or lack thereof, between mRNA sequencing and ribosome footprinting. As I have it now, the data I got on mRNA sequencing from another lab is reported in RNA counts - thus, to my understanding, there is no normalization to gene length. The data I have analyzed by ribosome footprinting, using cufflinks, is reported in FPKM. Thus, if I try to plot the ~200 genes for these 2 methods, there may be an inherent bias since FPKM is normalized to gene length. How can I run cufflinks to get an appropriate comparison (i.e. how do I get cufflinks to spit out RNA counts rather than FPKM)?
It might be not so easy to convert FPKM to read counts or vice-versa (it IS easy, but there are several small variations). In addition: 1) what did your colleague do when a read aligned on two different genes? 2) what if it aligned in 20 different places? 3) What if the read aligned in one place with 3 mismatches? 4) Did you use the same parameters as your colleagues? Therefore, I would suggest that you request the alignment files (or even better the read file) to your colleagues or at least you settle down for a standardized analysis.
Total guess but how about converting FPKM to TPM as described here and for the RNA counts dividing the RNA count by total number of RNAs in millon (N million RNAs sequenced) ? or just the RNA counts could be be used.