Learning the rules of peptide self-assembly through data mining with large language models

Abstract

Peptides are ubiquitous and important biomolecules that self-assemble into diverse structures. Although extensive research has explored the effects of chemical composition and exterior conditions on self-assembly, a systematic study consolidating these data to uncover global rules is lacking. In this work, we curate a peptide assembly database through a combination of manual processing by human experts and large language model–assisted literature mining. As a result, we collect over 1000 experimental data entries with information about peptide sequence, experimental conditions, and corresponding self-assembly phases. Using the data, machine learning models are developed, demonstrating excellent accuracy (>80%) in assembly phase classification. Moreover, we fine-tune a GPT model for peptide literature mining with the developed dataset, which markedly outperforms the pretrained model in extracting information from academic publications. This workflow can improve efficiency when exploring potential self-assembling peptide candidates, through guiding experimental work, while also deepening our understanding of the governing mechanisms.