This repository contains code used to build and interpret a deep learning model. It is a DNN classifier trained using gene expression data (TCGA). Then is interpreted to identify cancer specific gene expression signatures.
I was playing arund Deepcap website recently (BTW, very impressive work) and made input file like this (see file below). It's log normalized raw count matrix (I know it's not like what you recommanded, but only for a toy run). Columns are 901 genes from the 976 gene signatures. But no output nor error information available after few attempts. I wonder if it's because of the format or the missing 75 genes?