GamaPintoLab / GAP-MINE

GAP-MIME algorithm for gene functional annotation prediciton

GAP-MINE - Gene Annotation Prediction using Module-based Interpretable Network Embeddings

This project is part of the RNA Systems Biology Lab at BioISI - BioSystems and Integrative Sciences Institute. For other group projects, please visit our GitHub page.

Authors:

• Alexandre Coelho ([e-mail])
• Gonçalo Nogueira ([e-mail])
• João Miranda ([e-mail])
• Francisco Pinto ([e-mail])

For additional information, please contact Francisco Pinto [[e-mail]].

About

GAP-MINE is an algorithm for the prediction of gene associations with diseases and cell processes. This algorithm is composed of the following steps:

1. Graph Construction (Biological Network);
2. Proximity Metric Computation;
3. Additional Module Selection (Embedding Construction);
4. New Association Prediction.

Additionally, this repository also contains several notebooks for the visualization of the obtained results and the validation of the new predictions.

Technologies

• Python
• R
• jupyter
• Pandas
• NumPy
• SciPy
• scikit-learn
• OPLS-DA
• goatools

Notebooks

• Paper Figures
• Embedding Explainability
• GO and HPO Enrichment Analysis