Finding differential subpathways of metabolic networks using RNA-seq data

Question

[Benedek Dankó]

I would like to find a good approach or R/Python package that would help me to find differential subpathways of selected KEGG metabolic pathways. As input, I have expression data of N samples or differential expression data for G group comparisons. Let's say, I have differential expression data for cancer versus normal. Moreover, I also have KEGG gene sets in *.gmt files.

What kind of approach would you recommend to 1. map the differential expression or expression data to genes (nodes) of a selected metabolic network 2. find differential subpathways/node groups within a network between two sample groups?

So far I used MetaboSignal which is good for constricting the KEGG metabolic networks. I have also tried MIDAS to find differential subpathways, although it only worked with signaling pathways and not with metabolic pathways.

Any suggestions?

Answer 1

[alserg]

You can check out our gatom package: https://github.com/ctlab/gatom We're finishing it up to make more streamlined and accessible, so feedback is welcome. It is based on the general approach was developed a while, so you can read about it here (https://academic.oup.com/nar/article/44/W1/W194/2499310), but since then we've been continuously improving things. If you need any help setting it up, you can write to me directly to alsergbox at gmail dot com

[alserg]

Fyi, we've published the update, so you can read the description of the pipeline here: https://academic.oup.com/nar/advance-article/doi/10.1093/nar/gkac427/6594078