For networks pertaining to medicine, nodes represent biological factors (biomolecules, diseases, phenotypes, etc.)
It is possible, for example, to infer a bipartite graph representing the connections of diseases to their associated genes using the OMIM database.
Therefore, network medicine looks to identify the disease module for a specific pathophenotype using clustering algorithms.
[13] The topology of a biochemical reaction network determines the shape of drug dose-response curve[14] as well as the type of drug-drug interactions,[15] thus can help design efficient and safe therapeutic strategies.
In addition, the drug-target network (DTN) can play an important role in understanding the mechanisms of action of approved and experimental drugs.
[2] There can be a variety of ways to identifying drugs using network pharmacology; a simple example of this is the "guilt by association" method.
Such signaling modules are therapeutically best targeted at several sites, which is now the new and clinically applied definition of network pharmacology.
To achieve higher than current precision, patients must not be selected solely on descriptive phenotypes but also based on diagnostics that detect the module dysregulation.
[21] Epidemic models and concepts, such as spreading and contact tracing, have been adapted to be used in network analysis.
Bazzoni et al. (2015)[27] concluded that the DPNs of co-prescribed medications are dense, highly clustered, modular and assortative.
Askar et al. (2021)[28] created a network of the severe drug-drug interactions (DDIs) showing that it consisted of many clusters.
[31] It currently involves more than 80 Harvard Medical School (HMS) faculty and focuses on three areas: Massachusetts Institute of Technology offers an undergraduate course called "Network Medicine: Using Systems Biology and Signaling Networks to Create Novel Cancer Therapeutics".