Weighting interaction edges

Hi,

where can I find methods used for the representation of edges/relations in BiologicalNetworks? In other words how do you weight co-citation edges mentioning two genes or proteins in one Abstract/Sentence?
Also how do you integrate edges having different types of evidence, for example co-citation with protein-protein interaction and co-expression between the same two genes/proteins?
And the last, how do you derive causal/directional edges, do you do NLP of PubMed abstracts?

Thanks.

And Thank you for this piece of software, Good job!

Sebastian,

we use different types of edges: protein-protein, TF-DNA, Regulator-DNA, genetic, co-citation and other that were integrated from different data sources STRING, PID, HPRD, PubMed etc. The methods used in integration of all this data are described in the IntegromeDB publication in Bioinformatics and BiologicalNetworks publication in BMC Systems Biology (see the links on the front page)

Every edge property describes (see the Property Panel on the right) in details the type of interaction, data source, the score if applicable; for co-citation edge all publication mentioning two genes or proteins are provided. The Network Panel contains different types of filters where you can filter on edge scores, number of Abstracts supporting the evidence of interaction, etc.

Let us know if you have more questions.

Thanks.

Hello!