How can I assess the probability of a gene being disease-related?

Question

[gspirito]

Hello everyone, here's my problem:

I have a list of candidate genes and a list of genes related to a disease. I would like to know whether there is a tool that could help me assess the probability of each of my candidate genes of being related to that specific disease, integrating data from protein-protein interactions, co-expression and common ontologies.

I tried using utilities for building gene networks such as GeneMania or STRING, but those allow you to input one list of genes and I also found it difficult to understand whether there is some kind of p-value through which I could rank my candidate genes as more or less 'interconnected' with my disease-related genes.

Thank you in advance.

[Jean-Karim Heriche]

You can use GeneMania and related algorithms by using your disease genes as query and then looking for your candidate genes in the returned ranked list. If your candidate genes are ranked high then they are most likely related to the query (i.e. disease) genes. For an example of what I mean, see this paper.

Answer 1

[WouterDeCoster]

That process is called gene prioritization, and examples of tools for that would be ToppGene or ENDEAVOUR.