Bulk RNA-seq TPM values?

Technical
1

Hi team,

I downloaded two bulk RNA-seq datasets (H0351.2001, H0351.2002) from this link: Microarray Data Download :: Allen Brain Atlas: Human Brain .

I was planning to comparing to use the TPM values to check my gene of interest. However I noticed all values in “RNAseqTPM.csv” for each of the donor is extremely small (probably scaled by 10^(-6)) compare to other database (TPM values) for the same gene.

I’m curious how to interpret the numeric values in these RNAseqTPM.csv files. How the “TPM” is calculated? It didn’t look like a standard TPM values to me.

Thanks

Irene

2

Hi Irene,

TPM in this file is scaled so that the sum of the gene values per sample sum to one. When we used these values in our 2014 paper (Improving reliability and absolute quantification of human brain microarray data by filtering and scaling probes using RNA-Seq | BMC Genomics | Full Text), we converted the following way: datExprR = log2(1000000*datSeq + 1). I do not know what is considered best practice nowadays.

Best,

Jeremy