AmyloCore-ML: AI/Machine Learning Enabled Identification of Amyloid Fibril Core Regions Using Protein Language Models
- 1University of Denver
Abstract
Identification of amyloid forming regions in proteins remains central to understanding of their aggregation behaviour, which are central to various neurodegenerative and systemic disorders. Many predictors have been designed to identify amyloid forming regions, however they may not necessarily identify total residues found in actual disease associated fibril structures. To address this, we formed a sequence-based based predictor-AmyloCore-ML to ascertain structural amyloid-core propensity. The predictor was developed using experimentally determined fibril structures obtained from Amyloid Atlas. Ordered residues in the experimental structures were designated as core, and unresolved sampled from the same proteins as matched controls as non-core. Core and non-core regions were encoded using various physicochemical descriptors and protein language model (PLM) embeddings (ESM-2, ANKH, ProtT5) and evaluated using protein-grouped cross-validation. PLMs consistently outperformed physicochemical descriptors, with the best models reaching AUROC values of ~0.88 and AUPRC values of ~0.85. Locked full-length protein scans further localized experimental cores, with ESM-2/ExtraTrees achieving AUROC 0.833, AUPRC 0.751, and a mean peak distance of 12.4 residues. In comparative benchmarking, our predictor showed better performance metrics than CrossBeta and AggrescanAI. The framework therefore provides residue-resolved prediction of structurally incorporated amyloid-core regions directly from sequence.
Related
- The Identification of Biological Stains at Crime Scenes: A Promising Role for Proteomics and Machine Learning
- Identification of shared diagnostic biomarkers and immune-related mechanisms between sepsis and thoracic aortic aneurysm based on integrated bioinformatics and machine learning approaches
- Explainable computational framework for subtype classification and biomarker identification in head and neck squamous cell carcinoma using machine learning
- Advancing specific local wine identification through integration of ICP-MS mineral wine profiles and Machine Learning: A case study on discrimination among the ten Beaujolais crus
- Machine-learning-enhanced Optical Photothermal infrared (O-PTIR) spectroscopy for rapid simultaneous identification of multiple metabolites to monitor biotransformations: proof of concept on Brewer's spent grain (BSG)
BibTeX
@misc{singh2026amylocoreml,
title = {{AmyloCore-ML: AI/Machine Learning Enabled Identification of Amyloid Fibril Core Regions Using Protein Language Models}},
author = {J. Singh and R. Jangid and A. Srivastava},
howpublished = {bioRxiv},
year = {2026},
month = sep,
doi = {10.64898/2026.09.25.754455},
url = {https://www.biorxiv.org/content/10.64898/2026.09.25.754455v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - AmyloCore-ML: AI/Machine Learning Enabled Identification of Amyloid Fibril Core Regions Using Protein Language Models
AU - J. Singh
AU - R. Jangid
AU - A. Srivastava
PY - 2026
DA - 2026/09/25
PB - bioRxiv
DO - 10.64898/2026.09.25.754455
UR - https://www.biorxiv.org/content/10.64898/2026.09.25.754455v1
AB - Identification of amyloid forming regions in proteins remains central to understanding of their aggregation behaviour, which are central to various neurodegenerative and systemic disorders. Many predictors have been designed to identify amyloid forming regions, however they may not necessarily identify total residues found in actual disease associated fibril structures. To address this, we formed a sequence-based based predictor-AmyloCore-ML to ascertain structural amyloid-core propensity. The predictor was developed using experimentally determined fibril structures obtained from Amyloid Atlas. Ordered residues in the experimental structures were designated as core, and unresolved sampled from the same proteins as matched controls as non-core. Core and non-core regions were encoded using various physicochemical descriptors and protein language model (PLM) embeddings (ESM-2, ANKH, ProtT5) and evaluated using protein-grouped cross-validation. PLMs consistently outperformed physicochemical descriptors, with the best models reaching AUROC values of ~0.88 and AUPRC values of ~0.85. Locked full-length protein scans further localized experimental cores, with ESM-2/ExtraTrees achieving AUROC 0.833, AUPRC 0.751, and a mean peak distance of 12.4 residues. In comparative benchmarking, our predictor showed better performance metrics than CrossBeta and AggrescanAI. The framework therefore provides residue-resolved prediction of structurally incorporated amyloid-core regions directly from sequence.
ER -