Site4Drug: Predicting Drug-Binding Target Sites with an AI Agent

2026-06-01

AI Summary

This work addresses the challenges of ambiguous druggable site identification and high attrition rates in complex targets such as membrane proteins, which arise from limited accessibility, intricate topology, and post-translational modifications. To overcome these issues, we propose a multimodal perception-based AI agent that integrates diverse biological data—including protein sequence, 3D structure, topology, hydrophobicity, disulfide bonding, and domain context—without requiring prior assumptions about drug modality. The framework jointly infers optimal binding sites alongside their most suitable drug modalities, while harmonizing biophysical accessibility and chemical feasibility criteria to minimize false positives. The system generates an interpretable and traceable list of recommendations, complete with risk annotations and decision logs, thereby significantly enhancing the accuracy and practical utility of druggable region identification in structurally complex therapeutic targets.

Abstract

Selecting where to intervene on a protein (i.e., choosing a targetable site) is often a more ambiguous and failure-prone bottleneck than selecting what binds, especially for membrane proteins where accessibility, topology, and post-translational modifications (PTMs) constrain actionable regions. We present Site4Drug, a modality-aware site-finding agent that outputs a ranked list of targetable regions with explicit constraints, evidence summaries, risk flags, and a traceable decision log. Rather than requiring users to specify the drug modality upfront, Site4Drug can recommend a binding modality (e.g., antibody/peptide-like vs small-molecule) from the same evidence used for site discovery, including topology, hydropathy, PTM propensity, disulfides, domain context, and sequence. Importantly, this evidence is applied consistently across modalities, including small-molecule pocket discovery, to avoid selecting chemically plausible but biologically occluded sites.

Problem

• drug-binding site
• targetable site
• membrane protein
• post-translational modifications
• protein topology

Innovation

• AI agent
• drug-binding site prediction
• modality-aware
• targetable site selection
• biological constraints