search

DataX-JieHao / MiNet

Gene- and Pathway-based Framework for Multi-omics Data Intergration
MIT License
10 stars 2 forks source link

Input for MiNet #1

Open archanabhardwaj opened 4 years ago

archanabhardwaj commented 4 years ago

I really like your method. I would like to use it on my dataset. I have multi omics data in four different csv files. . From paper, its clear xi are generated as: pool all the feature in one matrix

Matrix (3x4) gene1_exp,gene1_meth,gene1_cnv,gene1_integration gene2_exp,gene2_meth,gene2_cnv,gene2_integration gene2_exp,gene2_meth,gene2_cnv,gene2_integration

Now as next step, How to create the "canonical gene expression" and pathway input layer for the MiNet model creation. Could you please elaborate it or provide the code ?

DataX-JieHao commented 4 years ago

I really like your method. I would like to use it on my dataset. I have multi omics data in four different csv files. . From paper, its clear xi are generated as: pool all the feature in one matrix

Matrix (3x4) gene1_exp,gene1_meth,gene1_cnv,gene1_integration gene2_exp,gene2_meth,gene2_cnv,gene2_integration gene2_exp,gene2_meth,gene2_cnv,gene2_integration

Now as next step, How to create the "canonical gene expression" and pathway input layer for the MiNet model creation. Could you please elaborate it or provide the code ?

Sorry for replying late. The canonical gene expression is controlled by the ordering of the input matrix and the mask between inputs and gene layer. So did the pathway layer.