search

vahuynh / GENIE3

Machine learning-based approach for the inference of gene regulatory networks from expression data.
68 stars 32 forks source link

format #12

Open fafuyyk opened 7 months ago

fafuyyk commented 7 months ago

Hello I am trying to use GENIE3 to infer the gene co-expression regulatory network, but I don’t quite understand some of the instructions in the tutorial, that is, the expression matrix cannot be normalized. Does that mean that the input result of the software can only be the original count instead of tpm or FPKM?

vahuynh commented 7 months ago

Hello,

The expression matrix can be normalized. What I meant in the tutorial is that the expression matrix does not necessarily need to be normalized if it is not necessary.

So, in the case of RNA-seq data, it is better to normalize the data to get the TPM or FPKM values, so that the expression values are comparable between genes and between samples. But once you have the TMP/FPKM values, you don't need to normalize any further (as opposed to other methods, where the data needs to be centered around zero, for example).