Modeling enzyme temperature stability from sequence segment perspective

2025-07-25

AI Summary

Enzyme thermal stability assessment via experimental assays is time-consuming, while existing computational models suffer from data scarcity and skewed distribution. To address these limitations, this work proposes the Segment Transformer framework—the first approach to model thermal stability at the protein sequence segment level, explicitly capturing region-specific contributions to stability. Evaluated on a high-quality, in-house temperature stability dataset, the model achieves RMSE = 24.03, MAE = 18.09, and Pearson correlation = 0.33, significantly outperforming baseline methods. Furthermore, guided by segment-level predictions, rational design of the lipase Cutinase yielded only 17 mutations, resulting in a 1.64-fold increase in residual activity after thermal treatment while fully preserving catalytic function. This demonstrates the practical utility of segment-level representations for enzyme directed evolution.

Technology Category

• Computer Vision: Segmentation
• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: Safety and Robustness

Application Category

• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Developing enzymes with desired thermal properties is crucial for a wide range of industrial and research applications, and determining temperature stability is an essential step in this process. Experimental determination of thermal parameters is labor-intensive, time-consuming, and costly. Moreover, existing computational approaches are often hindered by limited data availability and imbalanced distributions. To address these challenges, we introduce a curated temperature stability dataset designed for model development and benchmarking in enzyme thermal modeling. Leveraging this dataset, we present the extit{Segment Transformer}, a novel deep learning framework that enables efficient and accurate prediction of enzyme temperature stability. The model achieves state-of-the-art performance with an RMSE of 24.03, MAE of 18.09, and Pearson and Spearman correlations of 0.33, respectively. These results highlight the effectiveness of incorporating segment-level representations, grounded in the biological observation that different regions of a protein sequence contribute unequally to thermal behavior. As a proof of concept, we applied the Segment Transformer to guide the engineering of a cutinase enzyme. Experimental validation demonstrated a 1.64-fold improvement in relative activity following heat treatment, achieved through only 17 mutations and without compromising catalytic function.

Problem

• Predict enzyme temperature stability from sequence segments
• Overcome data limitations in computational thermal modeling
• Improve enzyme thermal stability via targeted mutations

Innovation

• Curated dataset for enzyme thermal modeling
• Segment Transformer deep learning framework
• Segment-level protein sequence representations