Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer

Membrane permeability is a critical bottleneck in the development of cyclic peptide drugs. Experimental membrane permeability testing is costly, and precise in silico prediction tools are scarce. In this study, we developed CPMP (https://github.com/panda1103/CPMP), a cyclic peptide membrane permeabi...

Full description

Saved in:
Bibliographic Details
Main Authors: Dawei Jiang, Zixi Chen, Hongli Du
Format: Article
Language:English
Published: Frontiers Media S.A. 2025-03-01
Series:Frontiers in Bioinformatics
Subjects:
Online Access:https://www.frontiersin.org/articles/10.3389/fbinf.2025.1566174/full
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850069465145802752
author Dawei Jiang
Zixi Chen
Zixi Chen
Hongli Du
author_facet Dawei Jiang
Zixi Chen
Zixi Chen
Hongli Du
author_sort Dawei Jiang
collection DOAJ
description Membrane permeability is a critical bottleneck in the development of cyclic peptide drugs. Experimental membrane permeability testing is costly, and precise in silico prediction tools are scarce. In this study, we developed CPMP (https://github.com/panda1103/CPMP), a cyclic peptide membrane permeability prediction model based on the Molecular Attention Transformer (MAT) frame. The model demonstrated robust predictive performance, achieving determination coefficients (R2) of 0.67 for PAMPA permeability prediction, and R2 values of 0.75, 0.62, and 0.73 for Caco-2, RRCK, and MDCK cell permeability predictions, respectively. Its performance outperforms traditional machine learning methods and graph-based neural network models. In ablation experiments, we validated the effectiveness of each component in the MAT architecture. Additionally, we analyzed the impact of data pre-training and cyclic peptide conformation optimization on model performance.
format Article
id doaj-art-2a12459e02024704a5db246a8fa2ed9d
institution DOAJ
issn 2673-7647
language English
publishDate 2025-03-01
publisher Frontiers Media S.A.
record_format Article
series Frontiers in Bioinformatics
spelling doaj-art-2a12459e02024704a5db246a8fa2ed9d2025-08-20T02:47:46ZengFrontiers Media S.A.Frontiers in Bioinformatics2673-76472025-03-01510.3389/fbinf.2025.15661741566174Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformerDawei Jiang0Zixi Chen1Zixi Chen2Hongli Du3School of Biology and Biological Engineering, South China University of Technology, Guangzhou, ChinaDepartment of Gerontology, ShenZhen Longhua District Central Hospital, Shenzhen, ChinaSchool of Safety Science and Engineering, Anhui University of Science and Technology, Huainan, ChinaSchool of Biology and Biological Engineering, South China University of Technology, Guangzhou, ChinaMembrane permeability is a critical bottleneck in the development of cyclic peptide drugs. Experimental membrane permeability testing is costly, and precise in silico prediction tools are scarce. In this study, we developed CPMP (https://github.com/panda1103/CPMP), a cyclic peptide membrane permeability prediction model based on the Molecular Attention Transformer (MAT) frame. The model demonstrated robust predictive performance, achieving determination coefficients (R2) of 0.67 for PAMPA permeability prediction, and R2 values of 0.75, 0.62, and 0.73 for Caco-2, RRCK, and MDCK cell permeability predictions, respectively. Its performance outperforms traditional machine learning methods and graph-based neural network models. In ablation experiments, we validated the effectiveness of each component in the MAT architecture. Additionally, we analyzed the impact of data pre-training and cyclic peptide conformation optimization on model performance.https://www.frontiersin.org/articles/10.3389/fbinf.2025.1566174/fullcyclic peptidemembrane permeabilitydeep learningmolecular attention transformerpampa
spellingShingle Dawei Jiang
Zixi Chen
Zixi Chen
Hongli Du
Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
Frontiers in Bioinformatics
cyclic peptide
membrane permeability
deep learning
molecular attention transformer
pampa
title Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
title_full Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
title_fullStr Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
title_full_unstemmed Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
title_short Cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
title_sort cyclic peptide membrane permeability prediction using deep learning model based on molecular attention transformer
topic cyclic peptide
membrane permeability
deep learning
molecular attention transformer
pampa
url https://www.frontiersin.org/articles/10.3389/fbinf.2025.1566174/full
work_keys_str_mv AT daweijiang cyclicpeptidemembranepermeabilitypredictionusingdeeplearningmodelbasedonmolecularattentiontransformer
AT zixichen cyclicpeptidemembranepermeabilitypredictionusingdeeplearningmodelbasedonmolecularattentiontransformer
AT zixichen cyclicpeptidemembranepermeabilitypredictionusingdeeplearningmodelbasedonmolecularattentiontransformer
AT honglidu cyclicpeptidemembranepermeabilitypredictionusingdeeplearningmodelbasedonmolecularattentiontransformer