AI-Driven De Novo Generation of Protein Binders
About the poster
This project aimed to leverage AI-driven methods, complemented by advanced computational techniques such as Molecular Modelling (MM) and Molecular Dynamics (MD) simulations, to generate novel binders for two prespecified protein targets: T1 and T2. The entire binder discovery process – from initial target analysis to final candidate selection – was performed in silico.
The core objective was to generate a ranked list of up to 100 binder candidates for each target through a comprehensive computational approach. This process encompasses the identification of promising binding sites, the design of novel binder structures and sequences using generative AI models, and rigorous in silico filtering and ranking based on predicted binding, stability, solubility, and other key biochemical properties.
Poster was presented at PEGS 2026 in Boston
