Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance

Understanding protein-protein interactions (PPIs) is essential because knowledge regarding PPIs helps in determining the biochemical functions of organisms. Advances made in machine learning techniques, for example, DeepFE-PPI, GcForest-PPI, and DeepPPI, have been applied to enhance PPI prediction p...

Full description

Saved in:
Bibliographic Details
Main Authors: Tran Hoai Nhan, Nguyen Phuc Xuan Quynh, Le Anh Phuong
Format: Article
Language:English
Published: Taylor & Francis Group 2025-04-01
Series:Journal of Information and Telecommunication
Subjects:
Online Access:https://www.tandfonline.com/doi/10.1080/24751839.2024.2411883
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849328995166846976
author Tran Hoai Nhan
Nguyen Phuc Xuan Quynh
Le Anh Phuong
author_facet Tran Hoai Nhan
Nguyen Phuc Xuan Quynh
Le Anh Phuong
author_sort Tran Hoai Nhan
collection DOAJ
description Understanding protein-protein interactions (PPIs) is essential because knowledge regarding PPIs helps in determining the biochemical functions of organisms. Advances made in machine learning techniques, for example, DeepFE-PPI, GcForest-PPI, and DeepPPI, have been applied to enhance PPI prediction performance. However, most methods use either handcrafted or learned features. Improving protein representation quality can significantly impact PPI model performance. This study aims to propose a deep neural network-based model, in which both handcrafted and learned features are fused to improve PPI prediction. In our method, the handcrafted features are computed by sequence descriptors, including Amino Acid Composition (AAC), Local Descriptors, Pseudo-AAC, Amphiphilic Pseudo-AAC, and Quasi-Sequence-Order. The learned features are generated by Word2vec after being trained on a large database of protein sequences. Standard evaluation methods and benchmark datasets are used to assess the proposed model. The results show that our model achieves 95.8% and 99.3% accuracy on the Saccharomyces cerevisiae and Human datasets, respectively. Additionally, the generalizability of our model is evaluated through independent tests, and our findings indicate that the proposed feature combination method correctly classifies experimentally validated interactions in eight datasets, including cross-species and PPI networks. The source code and data for this work are included at https://github.com/npxquynhdhsp/ComFeatPPI.git.
format Article
id doaj-art-141e769df7c14db4a5bdebbf58db0c9c
institution Kabale University
issn 2475-1839
2475-1847
language English
publishDate 2025-04-01
publisher Taylor & Francis Group
record_format Article
series Journal of Information and Telecommunication
spelling doaj-art-141e769df7c14db4a5bdebbf58db0c9c2025-08-20T03:47:24ZengTaylor & Francis GroupJournal of Information and Telecommunication2475-18392475-18472025-04-019215117210.1080/24751839.2024.2411883Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performanceTran Hoai Nhan0Nguyen Phuc Xuan Quynh1Le Anh Phuong2Department of Informatics, University of Education, Hue University, Hue city, VietnamDepartment of Informatics, University of Education, Hue University, Hue city, VietnamDepartment of Informatics, University of Education, Hue University, Hue city, VietnamUnderstanding protein-protein interactions (PPIs) is essential because knowledge regarding PPIs helps in determining the biochemical functions of organisms. Advances made in machine learning techniques, for example, DeepFE-PPI, GcForest-PPI, and DeepPPI, have been applied to enhance PPI prediction performance. However, most methods use either handcrafted or learned features. Improving protein representation quality can significantly impact PPI model performance. This study aims to propose a deep neural network-based model, in which both handcrafted and learned features are fused to improve PPI prediction. In our method, the handcrafted features are computed by sequence descriptors, including Amino Acid Composition (AAC), Local Descriptors, Pseudo-AAC, Amphiphilic Pseudo-AAC, and Quasi-Sequence-Order. The learned features are generated by Word2vec after being trained on a large database of protein sequences. Standard evaluation methods and benchmark datasets are used to assess the proposed model. The results show that our model achieves 95.8% and 99.3% accuracy on the Saccharomyces cerevisiae and Human datasets, respectively. Additionally, the generalizability of our model is evaluated through independent tests, and our findings indicate that the proposed feature combination method correctly classifies experimentally validated interactions in eight datasets, including cross-species and PPI networks. The source code and data for this work are included at https://github.com/npxquynhdhsp/ComFeatPPI.git.https://www.tandfonline.com/doi/10.1080/24751839.2024.2411883Sequence analysisprotein-protein interactionmachine learning
spellingShingle Tran Hoai Nhan
Nguyen Phuc Xuan Quynh
Le Anh Phuong
Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance
Journal of Information and Telecommunication
Sequence analysis
protein-protein interaction
machine learning
title Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance
title_full Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance
title_fullStr Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance
title_full_unstemmed Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance
title_short Combining handcrafted and learned features using deep learning to improve protein-protein interaction prediction performance
title_sort combining handcrafted and learned features using deep learning to improve protein protein interaction prediction performance
topic Sequence analysis
protein-protein interaction
machine learning
url https://www.tandfonline.com/doi/10.1080/24751839.2024.2411883
work_keys_str_mv AT tranhoainhan combininghandcraftedandlearnedfeaturesusingdeeplearningtoimproveproteinproteininteractionpredictionperformance
AT nguyenphucxuanquynh combininghandcraftedandlearnedfeaturesusingdeeplearningtoimproveproteinproteininteractionpredictionperformance
AT leanhphuong combininghandcraftedandlearnedfeaturesusingdeeplearningtoimproveproteinproteininteractionpredictionperformance