Predicting Multiple Types of Associations Between miRNAs and Diseases Based on Graph Regularized Weighted Tensor Decomposition

Abstract

Many studies have indicated miRNAs lead to the occurrence and development of diseases through a variety of underlying mechanisms. Meanwhile, computational models can save time, minimize cost, and discover potential associations on a large scale. However, most existing computational models based on a matrix or tensor decomposition cannot recover positive samples well. Moreover, the high noise of biological similarity networks and how to preserve these similarity relationships in low-dimensional space are also challenges. To this end, we propose a novel computational framework, called WeightTDAIGN, to identify potential multiple types of miRNA–disease associations. WeightTDAIGN can recover positive samples well and improve prediction performance by weighting positive samples. WeightTDAIGN integrates more auxiliary information related to miRNAs and diseases into the tensor decomposition framework, focuses on learning low-rank tensor space, and constrains projection matrices by using the L2,1 norm to reduce the impact of redundant information on the model. In addition, WeightTDAIGN can preserve the local structure information in the biological similarity network by introducing graph Laplacian regularization. Our experimental results show that the sparser datasets, the more satisfactory performance of WeightTDAIGN can be obtained. Also, the results of case studies further illustrate that WeightTDAIGN can accurately predict the associations of miRNA–disease-type.

A many‐objective optimization‐based local tensor factorization model for skin cancer detection

“…However, due to the lack of negative samples in medical data, false negative samples in the prediction results seriously affect the prediction accuracy. To address this issue, Ouyang et al 12,27 increased the weight of positive samples, introduced graph Laplacian regularization to preserve local biological similarity information, and combined it with tensor decomposition to improve the predictive performance.…”

miRGediNET: A comprehensive examination of common genes in miRNA-Target interactions and disease associations: Insights from a grouping-scoring-modeling approach

“…Furthermore, the current state of research is dispersed, with results frequently limited to particular diseases or miRNAs, which hinders our ability to create a comprehensive understanding of the interactions between miRNAs and diseases [ 3 ]. Additionally, there are insufficient computational tools to combine various biological data sources and present a complete picture of these interactions [ 4 ].…”

“…In [ 3 ], the authors introduced RBMMMDA, a method for predicting multiple types of disease-microRNA associations. RBMMMDA combines miRNA functional similarity, disease semantic similarity, and known miRNA-disease associations to create a weighted tensor representation.…”

Inferring human miRNA–disease associations via multiple kernel fusion on GCNII

“…proposed a signed graph neural network method (SGNNMD) to predict deregulation types of miRNA-disease associations. And WeightTDAIGN was proposed by ( Ouyang et al (2022) ) later. All these models are capable of identifying multiple types of miRNA-disease associations, but the performance of these models is not yet as good as that of those designed to identify single potential type of miRNA–disease association.…”