Fitness aligned structural modeling enables scalable virtual screening with AuroBind

2025-08-04

AI Summary

Over 96% of human proteins lack approved therapeutic agents, primarily due to the limited atomic-level accuracy and inability of conventional structure-based virtual screening methods to model ligand–protein binding adaptability. To address this, we introduce AuroBind—a unified framework integrating direct preference optimization, self-distillation of high-confidence protein–ligand complexes, and teacher–student acceleration—enabling joint prediction of binding poses and conformational adaptability. Fine-tuned on a million-compound chemogenomic dataset, AuroBind achieves atomic-resolution structural fidelity and functional awareness, enabling ultra-large-scale library screening with >10,000× speedup. In prospective screening across ten disease-relevant targets, experimental hit rates ranged from 7% to 69%, yielding sub-nanomolar to picomolar compounds. Notably, AuroBind identified first-in-class agonists and antagonists for orphan receptors GPR151 and GPR160, with functional modulation validated in hepatocellular and prostate cancer models.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Machine Learning: Bio-inspired Learning
• Humans and AI: Game Design — Virtual Humans, NPCs and Autonomous Characters

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Graph Algorithms and Modeling for the Web: Representation, reconstruction, and subgraph or motif discovery in Web-related graphs
• Economics, Online Markets and Human Computation: Sustainability of Web economics

Abstract

Most human proteins remain undrugged, over 96% of human proteins remain unexploited by approved therapeutics. While structure-based virtual screening promises to expand the druggable proteome, existing methods lack atomic-level precision and fail to predict binding fitness, limiting translational impact. We present AuroBind, a scalable virtual screening framework that fine-tunes a custom atomic-level structural model on million-scale chemogenomic data. AuroBind integrates direct preference optimization, self-distillation from high-confidence complexes, and a teacher-student acceleration strategy to jointly predict ligand-bound structures and binding fitness. The proposed models outperform state-of-the-art models on structural and functional benchmarks while enabling 100,000-fold faster screening across ultra-large compound libraries. In a prospective screen across ten disease-relevant targets, AuroBind achieved experimental hit rates of 7-69%, with top compounds reaching sub-nanomolar to picomolar potency. For the orphan GPCRs GPR151 and GPR160, AuroBind identified both agonists and antagonists with success rates of 16-30%, and functional assays confirmed GPR160 modulation in liver and prostate cancer models. AuroBind offers a generalizable framework for structure-function learning and high-throughput molecular screening, bridging the gap between structure prediction and therapeutic discovery.

Problem

• Addresses undrugged human proteins with scalable virtual screening
• Improves atomic-level precision in predicting binding fitness
• Enables rapid screening across ultra-large compound libraries

Innovation

• Fine-tunes atomic-level model on chemogenomic data
• Integrates preference optimization and self-distillation
• Enables ultra-fast screening with teacher-student strategy