Structure-Aware Multi-Label Therapeutic Peptide Predictor (SA-MTP), a structure-aware framework designed for multifunctional annotation of therapeutic peptides, is introduced.
Abstract
Abstract Therapeutic peptides show many biological activities and are now widely viewed as promising candidates for new drug development. Accurate functional annotation of therapeutic peptides is still difficult. This difficulty comes from their short sequence length, strong structural flexibility, and the presence of multiple biological functions within a single peptide.Here, we introduce Structure-Aware Multi-Label Therapeutic Peptide Predictor (SA-MTP), a structure-aware framework designed for multifunctional annotation of therapeutic peptides. SA-MTP combines pretrained protein language models with a graph attention network to capture sequence semantics and probabilistic structural features. Input-dependent structure-aware graphs are constructed to describe conformational variation, which is especially common in short peptides. Benchmarking experiments across 15 therapeutic function categories were conducted using datasets. The results show that SA-MTP achieves better performance than existing methods across several evaluation metrics, including accuracy, F1-score, and Matthews correlation coefficient.
PeptiVerse is a unified platform that leverages large foundation models to predict diverse peptide developability properties from both amino acid sequences and SMILES representations, enabling accessible, scalable analysis for peptide drug design.
A novel protein segment capture strategy for drug-target affinity prediction (SAPDTA), which is designed to extract local protein features through a local block capture approach, enabling more flexible extraction of protein structure information at different levels.
Zihao Fang, Guanqiu Qi, Stanley Tang et al.· Sensors and AI· 0 citations
Mapping the protein interactome is fundamental to understanding disease mechanisms and facilitating therapeutic development. Although protein language models (PLMs) such as ESM-2 have advanced protein-protein interaction (PPI) prediction, their high-dimensional representations remain difficult to connect to verifiable biological signals. To address this limitation, we propose HybridStack-PPI, a gray-box framework that combines ESM-2 sequence representations with explicit physicochemical and motif-derived biological descriptors. The architecture uses motif-anchored local pooling global mean pooling, symmetric pair encoding, fold-internal feature selection, LightGBM branch learners, and an elastic-net logistic-regression stacking layer. We evaluated the method using a C3 cluster-based cross-validation protocol with a 40% sequence-identity clustering threshold and a Same-GO hard-negative setting in which negative candidates shared functional annotations with positive pairs. Under this setting, HybridStack-PPI reached a Human ROC-AUC of 73.65%, PR-AUC of 91.35%, MCC of 28.06%, and specificity of 75.61%. The results indicate a conservative operating point: compared to more recall-oriented baselines, the proposed stack trades lower recall and F1 for higher specificity, MCC, and ranking behavior under functionally similar negative samples. We further reported cross-species transfer, ablation, latency, SHAP-based descriptor attribution, and meta-learner coefficient analyses to clarify both the promise and limitations of biologically informed PPI prediction.
T. T. Nguyen, X. Mai, N. Nguyen· IEEE Access· 0 citations
The results demonstrate the effectiveness of integrating multi-scale and multi-modal representations with cross-scale alignment for protein–RNA affinity prediction, and suggest that M2-PRNet can highlight relevant RNA-binding regions and support preliminary discrimination between strong and weak binders when plausible complex structures are available.
Junkai Wang, G. Luo, Yun-Song Yang et al.· Bioinformatics· 0 citations
Minimal Data Maximal Insight (MDMI), a two-stage structure-guided computational pipeline that designs functional peptide variants using only a small, annotated dataset, demonstrates that structure-informed pipelines can uncover remote functional sequence space from minimal data.
P. Bayat, Spencer J. Perkins, Sebastian Clancy et al.· bioRxiv· 0 citations