Aug 2026· Protein Science· Vol 35· 0 citations· 54 references
Medicine
TL;DR
Three modeling frameworks are developed, including models based on handcrafted features, models using embedding representations extracted from ProteinMPNN, and ensemble models integrating a diverse set of state‐of‐the‐art predictors integrating a diverse set of state‐of‐the‐art predictors.
Abstract
Accurately predicting mutation‐induced protein stability changes remains a central challenge in structural bioinformatics. Existing methods exhibit a strong bias toward destabilizing mutations, leading to limited performance for stabilizing mutations and constraining their utility in protein engineering. Here, we address this limitation through two complementary strategies: balanced dataset construction and integrative modeling. To mitigate the severe class imbalance in current stability datasets, we constructed undersampling‐based balanced datasets and further evaluated reverse‐mutation augmentation as a comparative strategy. Building on the rapid development of high‐performing predictors, we hypothesized that integrating their outputs could exploit complementary strengths and improve predictive accuracy. Accordingly, we developed three modeling frameworks, including models based on handcrafted features, models using embedding representations extracted from ProteinMPNN, and ensemble models integrating a diverse set of state‐of‐the‐art predictors. Across multiple independent test sets, ensemble models consistently outperformed individual approaches, with particularly pronounced gains in identifying stabilizing mutations. These findings demonstrate that combining undersampling‐based balanced data construction with systematic predictor integration provides an effective and practical strategy for achieving more balanced and accurate protein stability prediction, and offers a useful framework for identifying stabilizing mutations in protein engineering and related applications. StaMutAble is freely available at: https://github.com/minghuilab/StaMutAble.
UniStab is introduced, an end-to-end framework for predicting stability changes across all mutation types by leveraging the implicit geometric reasoning of a pre-trained folding model and demonstrates state-of-the-art performance, particularly in the challenging scenarios of multi-point mutations and indels.
Hong Tan, Shenggeng Lin, Yi Xiong· Chemical Science· 0 citations
Designing mutations that enhance protein stability is a central goal in protein engineering. However, experimentally screening large numbers of candidate mutations is costly and time-consuming, creating a strong need for computational methods that can identify potentially stabilizing mutations. Among these approaches, protein language models are particularly promising because they learn context-dependent amino acid preferences from large-scale sequence and structure datasets. Nevertheless, most existing stability prediction methods use these models primarily as feature extractors and do not fully exploit the amino acid probability distributions they encode. Here, we introduce MAXWELL (Matrix-wise Landscape Learning), a novel post-training method that calibrates the probabilistic outputs learned by protein language models during pretraining to generate mutational landscapes that quantify the effects of individual amino acid substitutions on protein stability. When applied to ProteinMPNN, MAXWELL yields a state-of-the-art predictor of the effects of protein mutations on stability, outperforming ThermoMPNN and other representative methods on a curated benchmark of experimentally measured stability changes. We next applied MAXWELL to the design of ten single-point mutations in the DhaA dehalogenase, seven of which (70%) increased thermal stability. Among them, G171W showed the largest improvement, with a measured ΔTm of 4.91 °C. These experimental results establish MAXWELL as a novel post-training strategy for protein language models and a practical framework for designing stabilizing mutations. Repository https://github.com/ai4protein/Venus-MAXWELL
Mingchen Li, Xiaoran Cheng, Fan Jiang et al.· bioRxiv· 0 citations
A systematic NMR-characterized dataset of mutants of the GA/GB model fold-switching system is presented and it is found that this benchmark revealed variable and position-dependent performance across methods, with certain AlphaFold2-based algorithms able to predict mutant effects at individual sites, indicating some understanding of physical effects of residue substitutions.
Nathaniel R. Felbinger, K. Carillo, Yihong Chen et al.· bioRxiv· 0 citations
Predicting protein stability, like changes in melting temperature (ΔTm) caused by mutations, is a critical task in therapeutic protein engineering and drug discovery. This is reflected by a growing solution space, including both AI-based sequence and structure based methods. This paper demonstrates that accurate ΔTm prediction does not require structural input features, but can achieve state-of-the-art results with a careful training design for large sequence-based protein language models. We combine an autoresearch-inspired setup search with controlled ablation studies and show that a well-tuned sequence-only ESM2-650M model [6] outperforms structure-informed methods in our benchmark, achieving the lowest error (MAE/RMSE) and competitive Pearson correlation without pH or structural inputs. We further show that choices such as loss function, pooling strategy, auxiliary supervision, and finetuning regime materially affect performance.
Daniel Siegismund, Mario Wieser, E. Natali et al.· bioRxiv· 0 citations
Coding mutations within intrinsically disordered regions (IDRs) of proteins are increasingly implicated in human diseases yet remain poorly interpreted by conventional variant-effect predictors that rely on structural stability and conservation-based metrics. Quantifying disruption of IDR-mediated liquid-liquid phase separation (LLPS) offers a biophysically principled approach to interpreting the pathogenic impact of such variants. However, existing LLPS predictors suffer from training biases toward self-separating proteins, show limited performance on partner- dependent phase separation, and often lack interpretability for variant prioritization. We present an interpretable ensemble machine-learning framework that integrates protein language model embeddings of sequence and predicted structure to predict LLPS propensity and classify proteins as self-separating or partner-dependent. Our two-step classifiers outperform existing methods on independent benchmark datasets, with the largest gains for partner-dependent LLPS proteins. Beyond classification, our framework identifies critical phase-separating regions and quantifies mutation-induced perturbations in LLPS. Applied to disease-associated variant databases, we found that pathogenic mutations are enriched in predicted phase-separating regions and frequently perturb LLPS propensity scores, implicating mutation-induced LLPS dysregulation as a potential pathogenic mechanism for numerous diseases. Overall, our framework provides an accurate, interpretable approach for identifying phase-separating proteins and linking aberrant phase- separation behavior to disease pathogenesis.
Protein language models have shown strong potential in modeling fitness landscapes for directed evolution; however, their predictive accuracy and generalization to unexplored sequence space remain limited under few-shot learning conditions. Here, we present MmALS (Multi-modal Active Learning System), a few-shot active learning framework that incorporates cold-start region scanning and multi-objective optimization to enable unbiased mutation-site selection. Its learning module employs a multimodal fusion architecture that incorporates sequence, structural, and multiple sequence alignment information, enabling accurate variant fitness prediction under low-sample conditions. The robustness of the MmALS logical framework was validated through in silico benchmarking across 12 independent datasets, in which it consistently accelerated convergence and demonstrated transferability across diverse enzyme fitness landscapes. Using Caldicellulosiruptor saccharolyticus Cellobiose 2-epimerase (CsCE) as the experimental validation, MmALS achieved a 3.88-fold enhancement of isomerization activity after only three iterative rounds (approximately 50 mutants per round), reaching a record-high activity of 20.72 U/mg. Overall, MmALS represents a data-efficient active learning framework for protein evolution, advancing computational enzyme design by enabling accurate optimization under limited experimental sampling.
Chenlu Zhu, Yucheng Guo, Yujie Yang et al.· ACS Catalysis· 0 citations