Improved Prediction of miRNA-Disease Associations Based on Matrix Completion with Network Regularization.

Chanyoung Park,Sanghyun Park,Chihyun Park,Jihwan Ha

doi:10.3390/cells9040881

Chanyoung Park, Sanghyun Park + Show 2 more

Open Access

https://doi.org/10.3390/cells9040881

Copy DOI

Journal: Cells	Publication Date: Apr 3, 2020
Citations: 26	License type: CC BY 4.0

Affiliation: University of Illinois Urbana-Champaign, Yonsei University

Abstract

The identification of potential microRNA (miRNA)-disease associations enables the elucidation of the pathogenesis of complex human diseases owing to the crucial role of miRNAs in various biologic processes and it yields insights into novel prognostic markers. In the consideration of the time and costs involved in wet experiments, computational models for finding novel miRNA-disease associations would be a great alternative. However, computational models, to date, are biased towards known miRNA-disease associations; this is not suitable for rare miRNAs (i.e., miRNAs with a few known disease associations) and uncommon diseases (i.e., diseases with a few known miRNA associations). This leads to poor prediction accuracies. The most straightforward way of improving the performance is by increasing the number of known miRNA-disease associations. However, due to lack of information, increasing attention has been paid to developing computational models that can handle insufficient data via a technical approach. In this paper, we present a general framework—improved prediction of miRNA-disease associations (IMDN)—based on matrix completion with network regularization to discover potential disease-related miRNAs. The success of adopting matrix factorization is demonstrated by its excellent performance in recommender systems. This approach considers a miRNA network as additional implicit feedback and makes predictions for disease associations relevant to a given miRNA based on its direct neighbors. Our experimental results demonstrate that IMDN achieved excellent performance with reliable area under the receiver operating characteristic (ROC) area under the curve (AUC) values of 0.9162 and 0.8965 in the frameworks of global and local leave-one-out cross-validations (LOOCV), respectively. Further, case studies demonstrated that our method can not only validate true miRNA-disease associations but also suggest novel disease-related miRNA candidates.

Highlights

MicroRNAs are small single-stranded non-coding RNAs that bind to the 30 untranslated regions (UTRs) of target messenger RNAs [1,2]. miRNAs tend to restrain gene expression by control of their own regulatory sequences and promoters; they bind to specific target mRNAs through base-paring, which inhibits the translation and stability
To demonstrate the superiority of improved prediction of miRNA-disease associations (IMDN), we compared our method with other state-of-the-art methods such as PMAMCA [23], matrix decomposition and heterogeneous graph inference (MDHGI) [28], ranking-based k-nearest neighbors for miRNA-disease association prediction (RKNNMDA) [26], random walk with restart for miRNA-disease association (RWRMDA) [15], MCMDA [24] and regularized least square for miRNA-disease association (RLSMDA) [21]
leave-one-out cross-validations (LOOCV) can be divided into global and local LOOCV, wherein each known miRNA-disease association was left out in turn as a test sample, whereas all the other remaining miRNA-disease pairs were considered as training samples

Summary

Introduction

MicroRNAs (miRNAs) are small single-stranded non-coding RNAs that bind to the 30 untranslated regions (UTRs) of target messenger RNAs (mRNAs) [1,2]. miRNAs tend to restrain gene expression by control of their own regulatory sequences and promoters; they bind to specific target mRNAs through base-paring, which inhibits the translation and stability. MiRNAs tend to restrain gene expression by control of their own regulatory sequences and promoters; they bind to specific target mRNAs through base-paring, which inhibits the translation and stability. Numerous studies continue to demonstrate the crucial roles of miRNAs in diverse biologic processes such as apoptosis [3], cell development [4], proliferation [5], viral infection [6] and metabolism [7]. Cells 2020, 9, 881 for diseases as well as potential prognostic biomarkers. Experiments further validated that miR-185 plays a crucial role in breast cancer by targeting Vegfa [11] and miR-122 inhibits cell proliferation and tumorigenesis of breast cancer by targeting IGF1R [12]. Predicting miRNA-disease associations can expand the understanding of molecular mechanisms of multiple human diseases and novel prognostic biomarkers

Objectives

Methods

Results

Discussion

Conclusion