Active learning of digenic functions with boolean matrix logic programming

2024/08/19 by Lun Ai, Stephen H. Muggleton, Ai, Lun +8 · 1 voice
Biochemistry, Genetics and Molecular Biology · Computer Science · Mathematics · #Arithmetic #Computer science #Mathematics #Programming language #Theoretical computer science #cs.AI #cs.LG #cs.SC #q-bio.MN #semigroups and automata theory

paper · pdf · doi:10.48550/arxiv.2408.14487

published in arXiv (Cornell University) (Cornell University)

openalex publication_date 2024/08/19 · arxiv published 2024/08/19 · openalex created_date 2024/09/21 · arxiv updated 2024/11/13 · openalex updated_date 2026/07/28

Abstract

We apply logic-based machine learning techniques to facilitate cellular engineering and drive biological discovery, based on comprehensive databases of metabolic processes called genome-scale metabolic network models (GEMs). Predicted host behaviours are not always correctly described by GEMs. Learning the intricate genetic interactions within GEMs presents computational and empirical challenges. To address these, we describe a novel approach called Boolean Matrix Logic Programming (BMLP) by leveraging boolean matrices to evaluate large logic programs. We introduce a new system, BMLP_active, which efficiently explores the genomic hypothesis space by guiding informative experimentation through active learning. In contrast to sub-symbolic methods, BMLP_active encodes a state-of-the-art GEM of a widely accepted bacterial host in an interpretable and logical representation using datalog logic programs. Notably, BMLP_active can successfully learn the interaction between a gene pair with fewer training examples than random experimentation, overcoming the increase in experimental design space. BMLP_active enables rapid optimisation of metabolic models and offers a realistic approach to a self-driving lab for microbial engineering.

Discussions

• Active learning of digenic functions with boolean matrix logic programming - https://arxiv.org/abs/2408.14487 [bsky, 0 points, 0 comments]

Related

• Boolean matrix logic programming for active learning of gene functions in genome-scale metabolic network models
• Human Comprehensible Active Learning of Genome-Scale Metabolic Networks
• BioReason: Incentivizing Multimodal Biological Reasoning within a DNA-LLM Model
• LGEM⁺: a first-order logic framework for automated improvement of metabolic network models through abduction
• Automated Biodesign Engineering by Abductive Meta-Interpretive Learning