Hi,
I got my RNA-Seq expression Data in FPKM, then I pick some interested genes to validate by qPCR. The expression table of FPKM and qPCR in like below:
There gene FPKM is 0 in sample B. How can I validate the expression the log2 ration of sampleA / SampleB.
I have tried to replace 0 to 0.0001, so that I can compare the log2 ration of RNA-Seq and qPCR and make a linear regression. But the r of regression is different when I replace 0 to another value.
Is there any good way to handle these data. These genes are in a pathway of my interested, I want to validate these genes expression in qPCR.
The problem here is how interpret these differences. I understand from your table that your RNA-Seq indicates that there is not expression of your genes, whereas the qPCR indicates that this not true.
What of these two pieces of information shall we trust?
In my opinion, RNA-Seq contains a trusted manner to quantify your expression. You map your sequences to the genome or transcriptome. If you do so with a 100 or 150b read, you get what is real. If you test RNA integrity and this is high before running the sequencing, I don't foresee further problems excepting if you are selecting your mRNA by their poly(A) tail and your RNA does not contain it
However, the number of potential pitfalls with the qPCR is huge in comparison. You, for example, don't exactly know what are being amplified, even though you run a melting analysis at the end of the run. You need to add many other putative artifacts, such as how well the primers align, whether you have contaminants in your samples or reagents, whether your reagents are in good conditions, etc
