Whether the probe intensity of microarray can be used to calculate TPM like the count data of RNAseq?

Question

[Nich]

Hello everyone！

I'm using a software that requires TPM. But I can't find enough RNAseq data to do analysis at the moment. I have downloaded some microarray data from a database that meets the analysis requirements. The original data were obtained using Affymetrix Human Genome U133A 2.0 Array.

Since I have not done microarray-related analysis before, I don't know about the relationship between microarray data and RNAseq data.

My question is that whether the probe intensity of microarray can be used to calculate TPM like the count data of RNAseq?

Thank you in advance for your answers!

[nicolas.goffin]

Hello, I have the same problem as you. Would it make sense to multiply the intensity values of the probes by a sufficiently large number (e.g. x 10000) and round up to get a sort of "count". I have no idea if this is possible or not...

Nicolas

[Gordon Smyth]

No, that would not make sense at all.

What analysis are you trying to do that requires TPMs?

Answer 1

[Gordon Smyth]

No, microarrays do not provide TPM values. They don't provide transcript expression levels in any linear sense. Affymetrix U113A arrays measure 3' expression using probes. They cannot distinguish transcripts for the same gene and the intensities returned depend on probe affinities as well as on gene expression.

Nevertheless there is a long history of successful and informative gene expression analyses using the Affymetric U133A arrays. What sort of analysis are you trying to do that can only be done on TPMs? I find it hard to think of any common gene-oriented expression analysis that truly requires TPMs.