How to link disease-associated genes and deferentially expressed genes in a protein-protein interaction network?

Question

[Ting-You Wang]

Hi, could anybody help me about how to link disease-associated genes and DEGs in a PPI network? and what's more, identify meaningful biological pathways or modules by using some approaches? the software/plugins and steps?

Here is my problem:

(1) a list of disease-associated genes from GWAS study and a list of genes which expressed deferentially between cases/controls, the two lists are quite different, indeed, which means there is nearly no overlap between the two lists.

(2) I throw them in a PPI network.

(3) how to link the two lists in a meaningful way and spawn some functional modules?

Thank you so much!

Answer 1

[Diwan]

One commonly used approach is to use pathway enrichment analysis (KEGG/Reactome). Sometimes, although there is no overlap in the gene lists they will map to the same pathway.

We can use PPI to generate de novo pathways (subnetworks). Check the cytoscape tool: KeyPathwayMiner. Link: http://keypathwayminer.mpi-inf.mpg.de/

Once you extract PPI subnetwork using DEGs, you can map the disease genes to the subnetwork. I found Keypathwayminer very useful. We have used it to extract large liver fibrosis subnetwork and followed it by clustering with clusterviz (another cytoscape tool) to get PPI modules. Check here: http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0112193