Looking at some publicly available RNAseq data. I ran cuffdiff on a pair of tumor-normal RNAseq data downloaded from TCGA, and I'm wondering what those results mean.
In the gene_exp.diff or the genes.fpkm_tracking file, if I add up all the FPKM values for both samples, one of the sums is 2.6 times as large as the other sum.
If I add up all the counts for both samples, one of the sums is 0.86 times as large as the other sum.
Why is there such a large difference in FPKM when read counts are comparable? You may say, well all the reads in one sample get mapped to large genes in one sample but small genes in the other sample..... but what does that imply?
Thanks.
In the gene_exp.diff or the genes.fpkm_tracking file, if I add up all the FPKM values for both samples, one of the sums is 2.6 times as large as the other sum.
If I add up all the counts for both samples, one of the sums is 0.86 times as large as the other sum.
Why is there such a large difference in FPKM when read counts are comparable? You may say, well all the reads in one sample get mapped to large genes in one sample but small genes in the other sample..... but what does that imply?
Thanks.