I-Corps: Translation Potential of a Computational Ecosystem for Protein Engineering

This I-Corps project focuses on the development of a computational ecosystem for intelligent protein engineering that enables the rapid and efficient design of enzyme variants for applications in biomedicine, biotechnology, and sustainability. The ability to engineer proteins with desired functional properties is essential for addressing challenges in pharmaceutical development, industrial biocatalysis, and environmental remediation.

However, traditional experimental approaches are costly and time-intensive, limiting the accessibility of enzyme engineering to a broader range of researchers. The technology provides a scalable, high-throughput solution that integrates molecular modeling with artificial intelligence to streamline protein design. By improving the efficiency of enzyme discovery, this technology enhances the ability to design novel routes for drug synthesis, improve diagnostic tools, and create environmentally friendly catalysts.

The adoption of this platform has the potential to significantly reduce industrial waste, lower energy consumption, and accelerate scientific progress in multiple industrial sectors.

This I-Corps project utilizes experiential learning coupled with a first-hand investigation of the industry ecosystem to assess the translation potential of the technology. This solution is based on the development of an integrated computational platform that combines quantum chemistry, molecular simulations, bioinformatics, and machine learning to predict the functional effects of enzyme mutations.

The approach leverages high-throughput molecular modeling to generate large datasets of molecular features that augment deep-learning algorithms, improving the accuracy of mutation effect predictions. Unlike traditional experimental screening methods, which rely on costly and time-consuming assays, this computational approach significantly reduces the time and resources required for enzyme optimization.

Complementary to existing machine learning models that are trained solely on enzyme sequences, these molecular features inform the structural basis underlying enzyme functions, thereby enhancing the interpretability of the model. The platform's predictive capabilities allow for the rational design of protein variants with enhanced activity, stability, and selectivity, broadening its applicability across pharmaceuticals, industrial enzyme manufacturing, and synthetic biology.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.