Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network

Abstract Background Studies have found that miRNAs play an important role in many biological activities involved in human diseases. Revealing the associations between miRNA and disease by biological experiments is time-consuming and expensive. The computational approaches provide a new alternative....

Full description

Bibliographic Details
Main Authors: Minghui Liu, Jingyi Yang, Jiacheng Wang, Lei Deng
Format: Article
Language:English
Published: BMC 2020-10-01
Series:BMC Medical Genomics
Subjects:
Online Access:http://link.springer.com/article/10.1186/s12920-020-00783-0
id doaj-32e6fbe5dacb4560b04332a5e6273eb3
record_format Article
spelling doaj-32e6fbe5dacb4560b04332a5e6273eb32021-04-02T12:53:02ZengBMCBMC Medical Genomics1755-87942020-10-0113S1011110.1186/s12920-020-00783-0Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous networkMinghui Liu0Jingyi Yang1Jiacheng Wang2Lei Deng3School of Computer Science and Engineering,Central South UniversitySchool of Computer Science and Engineering,Central South UniversitySchool of Computer Science and Engineering,Central South UniversitySchool of Computer Science and Engineering,Central South UniversityAbstract Background Studies have found that miRNAs play an important role in many biological activities involved in human diseases. Revealing the associations between miRNA and disease by biological experiments is time-consuming and expensive. The computational approaches provide a new alternative. However, because of the limited knowledge of the associations between miRNAs and diseases, it is difficult to support the prediction model effectively. Methods In this work, we propose a model to predict miRNA-disease associations, MDAPCOM, in which protein information associated with miRNAs and diseases is introduced to build a global miRNA-protein-disease network. Subsequently, diffusion features and HeteSim features, extracted from the global network, are combined to train the prediction model by eXtreme Gradient Boosting (XGBoost). Results The MDAPCOM model achieves AUC of 0.991 based on 10-fold cross-validation, which is significantly better than that of other two state-of-the-art methods RWRMDA and PRINCE. Furthermore, the model performs well on three unbalanced data sets. Conclusions The results suggest that the information behind proteins associated with miRNAs and diseases is crucial to the prediction of the associations between miRNAs and diseases, and the hybrid feature representation in the heterogeneous network is very effective for improving predictive performance.http://link.springer.com/article/10.1186/s12920-020-00783-0miRNA-disease associationHeteSim measureDiffusion featureeXtreme gradient boosting
collection DOAJ
language English
format Article
sources DOAJ
author Minghui Liu
Jingyi Yang
Jiacheng Wang
Lei Deng
spellingShingle Minghui Liu
Jingyi Yang
Jiacheng Wang
Lei Deng
Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network
BMC Medical Genomics
miRNA-disease association
HeteSim measure
Diffusion feature
eXtreme gradient boosting
author_facet Minghui Liu
Jingyi Yang
Jiacheng Wang
Lei Deng
author_sort Minghui Liu
title Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network
title_short Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network
title_full Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network
title_fullStr Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network
title_full_unstemmed Predicting miRNA-disease associations using a hybrid feature representation in the heterogeneous network
title_sort predicting mirna-disease associations using a hybrid feature representation in the heterogeneous network
publisher BMC
series BMC Medical Genomics
issn 1755-8794
publishDate 2020-10-01
description Abstract Background Studies have found that miRNAs play an important role in many biological activities involved in human diseases. Revealing the associations between miRNA and disease by biological experiments is time-consuming and expensive. The computational approaches provide a new alternative. However, because of the limited knowledge of the associations between miRNAs and diseases, it is difficult to support the prediction model effectively. Methods In this work, we propose a model to predict miRNA-disease associations, MDAPCOM, in which protein information associated with miRNAs and diseases is introduced to build a global miRNA-protein-disease network. Subsequently, diffusion features and HeteSim features, extracted from the global network, are combined to train the prediction model by eXtreme Gradient Boosting (XGBoost). Results The MDAPCOM model achieves AUC of 0.991 based on 10-fold cross-validation, which is significantly better than that of other two state-of-the-art methods RWRMDA and PRINCE. Furthermore, the model performs well on three unbalanced data sets. Conclusions The results suggest that the information behind proteins associated with miRNAs and diseases is crucial to the prediction of the associations between miRNAs and diseases, and the hybrid feature representation in the heterogeneous network is very effective for improving predictive performance.
topic miRNA-disease association
HeteSim measure
Diffusion feature
eXtreme gradient boosting
url http://link.springer.com/article/10.1186/s12920-020-00783-0
work_keys_str_mv AT minghuiliu predictingmirnadiseaseassociationsusingahybridfeaturerepresentationintheheterogeneousnetwork
AT jingyiyang predictingmirnadiseaseassociationsusingahybridfeaturerepresentationintheheterogeneousnetwork
AT jiachengwang predictingmirnadiseaseassociationsusingahybridfeaturerepresentationintheheterogeneousnetwork
AT leideng predictingmirnadiseaseassociationsusingahybridfeaturerepresentationintheheterogeneousnetwork
_version_ 1721567289023660032