Hi all,
I've been doing some followup PCR to confirm inferences of differentially expressed genes identified in an RNAseq experiment. We have observed general consistency in terms of the direction of differential expression, but we have been surprised to observe that the qPCR estimates generally find orders of magnitude larger fold changes than observed in the RNAseq data. The Ct values are reasonable and the melting curves look great.
In a past life when I was doing a lot of expression microarray experiments, I found that this phenomenon occurred a lot- basically that the microarray systematically underestimated gene expression changes. This is the first time that I have done this sort of followup for an RNAseq experiment, however, and I was wondering if others had any intuition about what might be going on. From first principles I can't think of a reason that RNAseq and qPCR fold change estimates would be far off, and as I say the directionality os consistent. I am wondering if others with more experience doing this can lend a bit of intuition about what might be going on.
In case it helps, I'm modeling read counts directly with a nb-distributed glm, and estimating fold change from the coefficient estimates.
I've been doing some followup PCR to confirm inferences of differentially expressed genes identified in an RNAseq experiment. We have observed general consistency in terms of the direction of differential expression, but we have been surprised to observe that the qPCR estimates generally find orders of magnitude larger fold changes than observed in the RNAseq data. The Ct values are reasonable and the melting curves look great.
In a past life when I was doing a lot of expression microarray experiments, I found that this phenomenon occurred a lot- basically that the microarray systematically underestimated gene expression changes. This is the first time that I have done this sort of followup for an RNAseq experiment, however, and I was wondering if others had any intuition about what might be going on. From first principles I can't think of a reason that RNAseq and qPCR fold change estimates would be far off, and as I say the directionality os consistent. I am wondering if others with more experience doing this can lend a bit of intuition about what might be going on.
In case it helps, I'm modeling read counts directly with a nb-distributed glm, and estimating fold change from the coefficient estimates.
Comment