Expert-guided Bayesian optimization for sustainable protein formulation

Anna Thomas, Georgios Zaverdinos, Petros Mandalis, Andreas Orfanoudakis, Akihiro Takino, Sohum Patnaik, David J. Kraus, Panos Kostopoulos, Caroline Cotto, Nikos Tsiaparas, Dan Jurafsky, Aadit Patel

Track: Track 1: Original Research/Position/Education/Attention Track

Keywords: Bayesian optimization, LLM-guided optimization, variable selection, sustainable proteins, plant-based food formulation, high-dimensional optimization, sparse optimization, closed-loop scientific discovery

TL;DR: Expert-Guided BO: an LLM or human picks a low-dim subspace; BO runs inside it. Outperforms baselines on FormulateBench (24 plant-protein tasks). Deployed with a food company, EGBO outperformed a professional food scientist on plant-based yogurt.

Abstract:

Autonomous scientific discovery systems increasingly use LLMs to narrow design spaces before experiments are run, but this practice is double-edged: when the LLM is right, sample efficiency can improve dramatically; when it is wrong, the system can underperform random search. We formalize Expert-Guided Bayesian Optimization (EGBO), in which an expert, e.g. a human or LLM, selects a low-dimensional subspace for BO and may adaptively expand it over time.
We decompose EGBO's suboptimality into a selection gap and an optimization gap, and characterize the coverage–dimension tradeoff governing when expert guidance helps.
To support in silico prototyping before costly real-world deployment, we introduce FormulateBench, a suite of 24 plant-based formulation tasks, on which LLM-guided EGBO outperforms all tested baselines. When deployed to optimize two plant-based dairy products, EGBO improves utility, as assessed by a trained human panel, by 29\% and 26\% in 10 iterations each. In a comparison with a professional human food scientist given the same time budget, EGBO achieved near-perfect utility of 0.992, vs. 0.850 for the food scientist.

Submission Number: 121