Understanding the stability of a drug product is tremendously important for protecting patient safety and developing robust products. An ability to predict early in the development stage with quantitative accuracy what may happen during the shelf  life of a given pharmaceutical product is valuable, and may significantly affect product development strategy. In this framework we are developing a fully automated tool for API degradation prediction. This effort involves identification of API sites likely to undergo oxidation, and implementing on-the-fly quantum chemical kinetic rate calculations. This software tool will be capable of self-learning, as previously calculated rates are used as training data for future estimations.

This project is done in collaboration with and is sponsored by Pfizer Inc.