2
$\begingroup$

I know that there are different types of biological networks. Protein-protein interaction, gene interaction, metabolic networks, gene expression networks, etc. Each of these networks can be represented as a graph.

1) Is there a principled way to integrate these graphs into a single large graph? For example, probably most of these networks have nodes that are genes, or gene related things, and this could allow us to match nodes in one network to nodes in another network.

2) Is there a principled way to ask joint questions about these networks?

$\endgroup$
1
  • $\begingroup$ Hi Mendel, welcome to Bioinformatics StackExchange. If you want more informative answers, it would be great if you could add some more context around your question (i.e. what particular problem are you trying to solve; be as specific as possible). There are a lot of different ways to approach bioinformatics problems, and this leads to a lot of garden-path discussion that doesn't really end up helping the person who has asked the question. $\endgroup$
    – gringer
    Feb 6 '19 at 20:31
2
$\begingroup$

I believe you're referring to something related to multi-layer networks. Here you can find an interesting reading about this kind of networks:

https://academic.oup.com/comnet/article/2/3/203/2841130

Also you can try with this one:

https://www.sciencedirect.com/science/article/pii/S0370157314002105

As a matter of fact, I'm not an expert of multi-layer networks but, as far as I can tell, the main problem (as you mentioned) is to find correlations, or connections, between the different layers in order to be able to draw some meaningful conclusion.

I am not sure you can call this a multi-layer network but this can be an example of what you're looking for: suppose you have a protein-protein interaction network. You can enrich your network using biological information. You can associate biological functions, or cellular locations or whatever you can find in GeneOntology (just to make an example of a biological database), as a further layer of information regarding your protein-protein interaction network.

This will help you in elucidating the role a node (or more nodes) is playing with respect to its neighbours: let's say five nodes form a clique and you also know they share the same cellular location...then you're able to say something about these nodes (maybe they form a complex?) that has a dual nature: biological and graph-based.

Hope it helps!

$\endgroup$
1
$\begingroup$

You are talking network statistics and this is a complex area of stats.

For generalised simplicity, Google fusion tables have a simple to use but impressively interactive network tool. This is a great starting point for exploration. Have a go and see what you think.

If you are looking at a haplotype map, there is a formal solution to this (parsimony network), which is standalone program written some time ago but still useful today. In my area split-decomposition is an important important technique.

The easiest question is the number of connecting nodes and Google fusion tables is as good as any routine interactive networking tool. As I say beyond this network stats is a specialised field.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.