Can Graph Theory Help Predict Cancer Progression?

[gravenewworld]

Let's say I had a network of enzymes that are all interconnected that may be involved in cancer progression. Each enzyme produces a chemical product that might be used by some other member in this network, but each enzyme might produce a product at different rates. Is there a way I could possibly use graph theory to model this network, along with the rate traffic through this enzyme network, in order to make some predictions on the final "product" of this network (the final product of which assists in cancer)?

I've had some graph theory before, but is there some way to incorporate the "rate of traffic" parameter into such a graph? So not just figuring the number of possible ways it might be possible to synthesize a final product, but how much and how fast we expect it to happen? What are some topics I can look up to point me in the right direction with regards to graph theory?

 

[mfb]

Can you model "rate of traffic" as weight for the edges?
Time-dependence might be tricky, unless you include some additional parameters for the edges.

It is possible to model your network as graph, the question is how much graph theory do you want to apply to it ;).

 

[Jack21222]

There's a whole field dedicated to this sort of thing; network science. They combine graph theory and some other disciplines to solve problems exactly like you describe.

So, you don't need to reinvent the wheel here. Check out some of the papers on Barabasi's website (I think it's www.barabasilab.com). You might find papers that have already done what you propose.

EDIT: Here is one paper that kind of scratches the surface and might be a decent starting point: http://jeb.biologists.org/content/210/9/1548.short

 

[gravenewworld]

Hmm thanks for the responses and paper leads. I realize there's a whole area in systems biology dedicated to this sort of thing, but was wondering if someone had a lead on a paper like the one you posted that would save me time on where to start. Quite an interesting read.

Been going through a sort of identity crisis lately. Am I a chemist? No. Am I a biologist? No. Am I an engineer/mathematician? No. I'm basically a guy that knows how to do some things from all of those fields. It would be neat to models something I'm studying mathematically to make predictions, test it biologically, and then pharmacologically treat a suspected network with a smarter designed molecule.

Cheers.