mims-harvard / pathways

Disease pathways in the human interactome

Pathways

We perform a large-scale analysis of disease pathways in the human interactome to better understand connectivity and higher-order network structure of disease pathways.

Discovering disease pathways, which can be defined as sets of proteins associated with a given disease, is an important problem that has the potential to provide clinically actionable insights for disease diagnosis, prognosis, and treatment.

Here we study the PPI network structure of disease pathways. We find that 90% of pathways do not correspond to single well-connected components in the PPI network. Instead, proteins associated with a single disease tend to form many separate connected components/regions in the network. We then evaluate state-of-the-art disease pathway discovery methods and show that their performance is especially poor on diseases with disconnected pathways.

We conclude that network connectivity structure alone may not be sufficient for disease pathway discovery. However, we show that higher-order network structures, such as small subgraphs of the pathway, provide a promising direction for the development of new methods.

Please check the project website for more details.

Citing

If you find disease pathway analysis useful for your research, please consider citing:

@inproceedings{agrawal2018,
  author={Agrawal, Monica and Zitnik, Marinka and Leskovec, Jure},
  title = {Large-scale Analysis of Disease Pathways in the Human Interactome},
  year = {2018},
  booktitle = {Pacific Symposium on Biocomputing},
  volume = {23},
  pages = {111-122}
}

Miscellaneous

Please send any questions you might have about the code and/or the algorithm to [email protected].

License

Decagon is licensed under the MIT License.

pathways's Issues

there seems mistakes in the script of communicty score calculation

Do you mean adjacency_matrix = nx.adjacency_matrix(nx_graph)?

Another place, in the final line, Do you mean disease_properties["normalized_cut"] = disease_properties["conductance"] + external_edges/(2*(num_total_edges-num_internal_edges)+external_edges)?