Heterogeneous Types of miRNA-Disease Associations Stratified by Multi-Layer Network Embedding and Prediction.

Yu, D-L; Yu, Z-G; Han, G-S; Li, J; Anh, V

Heterogeneous Types of miRNA-Disease Associations Stratified by Multi-Layer Network Embedding and Prediction.

Yu, D-L Yu, Z-G Han, G-S Li, J

Anh, V

Permalink

Publisher:: MDPI
Publication Type:: Journal Article
Citation:: Biomedicines, 2021, 9, (9), pp. 1-14
Issue Date:: 2021-09-03

Open Access

Copyright Clearance Process

Recently Added
In Progress
Open Access

This item is open access.

Adobe PDF

Download Published versionAdobe PDF (420.61 kB)

View on publisher's site

View statistics

Full metadata record

Field	Value	Language
dc.contributor.author	Yu, D-L
dc.contributor.author	Yu, Z-G
dc.contributor.author	Han, G-S
dc.contributor.author	Li, J https://orcid.org/0000-0003-1833-7413
dc.contributor.author	Anh, V
dc.date.accessioned	2022-05-14T03:42:05Z
dc.date.available	2021-08-30
dc.date.available	2022-05-14T03:42:05Z
dc.date.issued	2021-09-03
dc.identifier.citation	Biomedicines, 2021, 9, (9), pp. 1-14
dc.identifier.issn	2227-9059
dc.identifier.issn	2227-9059
dc.identifier.uri	http://hdl.handle.net/10453/157339
dc.description.abstract	Abnormal miRNA functions are widely involved in many diseases recorded in the database of experimentally supported human miRNA-disease associations (HMDD). Some of the associations are complicated: There can be up to five heterogeneous association types of miRNA with the same disease, including genetics type, epigenetics type, circulating miRNAs type, miRNA tissue expression type and miRNA-target interaction type. When one type of association is known for an miRNA-disease pair, it is important to predict any other types of the association for a better understanding of the disease mechanism. It is even more important to reveal associations for currently unassociated miRNAs and diseases. Methods have been recently proposed to make predictions on the association types of miRNA-disease pairs through restricted Boltzman machines, label propagation theories and tensor completion algorithms. None of them has exploited the non-linear characteristics in the miRNA-disease association network to improve the performance. We propose to use attributed multi-layer heterogeneous network embedding to learn the latent representations of miRNAs and diseases from each association type and then to predict the existence of the association type for all the miRNA-disease pairs. The performance of our method is compared with two newest methods via 10-fold cross-validation on the database HMDD v3.2 to demonstrate the superior prediction achieved by our method under different settings. Moreover, our real predictions made beyond the HMDD database can be all validated by NCBI literatures, confirming that our method is capable of accurately predicting new associations of miRNAs with diseases and their association types as well.
dc.format	Electronic
dc.language	eng
dc.publisher	MDPI
dc.relation.ispartof	Biomedicines
dc.relation.isbasedon	10.3390/biomedicines9091152
dc.rights	info:eu-repo/semantics/openAccess
dc.title	Heterogeneous Types of miRNA-Disease Associations Stratified by Multi-Layer Network Embedding and Prediction.
dc.type	Journal Article
utslib.citation.volume	9
utslib.location.activity	Switzerland
pubs.organisational-group	/University of Technology Sydney
pubs.organisational-group	/University of Technology Sydney/Faculty of Engineering and Information Technology
pubs.organisational-group	/University of Technology Sydney/Strength - CHT - Health Technologies
pubs.organisational-group	/University of Technology Sydney/Strength - AAI - Advanced Analytics Institute Research Centre
pubs.organisational-group	/University of Technology Sydney/Centre for Health Technologies (CHT)
utslib.copyright.status	open_access	*
pubs.consider-herdc	false
dc.date.updated	2022-05-14T03:42:03Z
pubs.issue	9
pubs.publication-status	Published online
pubs.volume	9
utslib.citation.issue	9

Abstract:

Abnormal miRNA functions are widely involved in many diseases recorded in the database of experimentally supported human miRNA-disease associations (HMDD). Some of the associations are complicated: There can be up to five heterogeneous association types of miRNA with the same disease, including genetics type, epigenetics type, circulating miRNAs type, miRNA tissue expression type and miRNA-target interaction type. When one type of association is known for an miRNA-disease pair, it is important to predict any other types of the association for a better understanding of the disease mechanism. It is even more important to reveal associations for currently unassociated miRNAs and diseases. Methods have been recently proposed to make predictions on the association types of miRNA-disease pairs through restricted Boltzman machines, label propagation theories and tensor completion algorithms. None of them has exploited the non-linear characteristics in the miRNA-disease association network to improve the performance. We propose to use attributed multi-layer heterogeneous network embedding to learn the latent representations of miRNAs and diseases from each association type and then to predict the existence of the association type for all the miRNA-disease pairs. The performance of our method is compared with two newest methods via 10-fold cross-validation on the database HMDD v3.2 to demonstrate the superior prediction achieved by our method under different settings. Moreover, our real predictions made beyond the HMDD database can be all validated by NCBI literatures, confirming that our method is capable of accurately predicting new associations of miRNAs with diseases and their association types as well.

Please use this identifier to cite or link to this item:

http://hdl.handle.net/10453/157339