This project utilizes the powers of various machine learning models to predict whether a pre-approved drug is effective against SARS-CoV-2 and whether that molecule is phospholipidosis-inducing, the phenomenon that inhibits a drug from in vitro to in vivo translation. The knowledge gained from this project can provide new opportunities to screen larger libraries in a context closer to in vitro and thus judge additional already-approved drugs for repurposing against COVID-19 and its variants.

This GitHub page demonstrates the development of this project from the initial proposal, explorative data analysis and visualization, model training, statistic evaluation on performance, to the final results and presentation. We hope our project can inspire future researchers to dig in more and explore additional exciting possibilities in this interdisciplinary field between pharmacology and machine learning.