AlloGen: Conformation-Selective Binder Generation with Differential State Scoring

2026-06-03

AI Summary

Traditional protein binder design often focuses solely on optimizing affinity while neglecting selective recognition of specific conformational states of target proteins—such as active or inactive states of kinases, GPCRs, or nuclear receptors—resulting in insufficient functional specificity. This work proposes AlloGen, a framework that, for the first time, formulates conformational selectivity as a learnable property. It introduces a generator-agnostic, fully differentiable, SE(3)-invariant graph Transformer scorer \( Q_\theta \), enabling gradient-guided optimization or reranking strategies without retraining, and thus compatible with any scaffold generator. Through a two-stage curriculum learning approach, the model jointly optimizes interface geometry and conformational discrimination. On benchmarks spanning multiple protein families, AlloGen consistently generates binders favoring the target conformation; experimental validation demonstrates that designed calmodulin-binding peptides selectively engage only the holo state, showing no detectable binding to the apo state.

Abstract

Protein binder design has largely optimized for affinity alone, leaving conformational selectivity unaddressed: for allosteric targets such as kinases, nuclear receptors, and GPCRs, a binder that engages both active and inactive states provides no functional specificity regardless of how tightly it binds. We introduce AlloGen, a modular framework that decouples backbone generation from a learned state-selectivity scorer $Q_θ$, an SE(3)-invariant interface graph transformer trained via a two-phase curriculum that first learns interface geometry before imposing conformational discrimination. Because $Q_θ$ is fully differentiable and generator-agnostic, it integrates with any backbone generator as a passive reranker or an active gradient-based guide without retraining. Across a diverse benchmark of proteins spanning multiple families and conformational mechanisms, AlloGen consistently identifies binders that preferentially recognize desired structural states while rejecting alternative conformations. Experimental validation on calmodulin further demonstrates that these computational selectivity signals translate to physical molecules, yielding de novo peptides that bind the desired holo conformation while exhibiting no detectable binding to the apo state. Together, these results establish conformational selectivity as a learnable property and provide a general framework for state-selective protein binder design.

Problem

• conformational selectivity
• protein binder design
• allosteric targets
• structural states
• functional specificity

Innovation

• conformational selectivity
• protein binder design
• SE(3)-invariant graph transformer
• differentiable scoring
• allosteric targeting