Implicit Generative Property Enhancer

Pedro O. Pinheiro, Pan Kessel, Aya Abdelsalam Ismail, Sai Pooja Mahajan, Kyunghyun Cho, Saeed Saremi, Natasa Tagasovska

Keywords: generative modeling, drug discovery, design optimization

TL;DR: We cast the problem of design optimization as as (implicit) conditional generation, we provide general theoretical guarantees and show its benefits empirically on standard protein engineering benchmarks

Abstract: Generative modeling is increasingly important for data-driven computational design. Conventional approaches pair a generative model with a discriminative model to select or guide samples toward optimized designs. Yet discriminative models often struggle in data-scarce settings, common in scientific applications, and are unreliable in the tails of the distribution where optimal designs typically lie. We introduce generative property enhancer (GPE), an approach that implicitly guides generation by matching samples with lower property values to higher-value ones. Formulated as conditional density estimation, our framework defines a target distribution with improved properties, compelling the generative model to produce enhanced, diverse designs without auxiliary predictors. GPE is simple, scalable, end-to-end, modality-agnostic, and integrates seamlessly with diverse generative model architectures and losses. We demonstrate competitive empirical results on standard _in silico_ offline (non-sequential) protein fitness optimization benchmarks. Finally, we propose iterative training on a combination of limited real data and self-generated synthetic data, enabling extrapolation beyond the original property ranges.

Primary Area: Machine learning for sciences (e.g. climate, health, life sciences, physics, social sciences)

Submission Number: 10992