How I Predicted the Effect of Mutations on Protein Interactions Using AlphaFold

Using AlphaFold-Multimer, XGBoost, and 47,000 SLURM jobs to predict PPI outcomes with 91% AUC

By Author

The human interactome (all protein-protein interactions) may number up to 600,000 interactions.

With so many possible protein-protein interactions (PPIs), predicting how a disease-causing mutation affects the interactome seems like a Herculean task — but not as impossible as you might expect.

(Especially when you give a University of Waterloo co-op student free access to a beefy GPU cluster, world-class mentorship, and free agency to pursue any approach).

Using the machine learning framework XGBoost, cutting-edge deep learning software AlphaFold-Multimer (AF-M), and over 47,000 SLURM jobs, I built a multi-classifier model that predicts the effects of missense mutations on PPIs with a 91% AUC.

Multi-class ROC curve and AUC (By Author)

In this article, I'm going to walk through:

• The Background: The research question and why we chose it.
• Data Acquisition & Processing: How and why we…