Research Of Calculating Methods Of MicroRNA Functional Similarity Based On Disease Association

MicroRNAs(miRNAs) are one important class of gene regulators,which play critical roles in a variety of biological processes,such as cell development,differentiation,proliferation,apoptosis,virus defense,disease formation,and development.Recently,miRNAs have drawn more attention from researchers.Theoretically,molecules are not completely independent,and they have various direct or indirect relationships with other molecules.Moreover,there is increasing evidence which demonstrates that miRNAs' regulatory functions are closely tied to a variety of human diseases.And the dysregulation of a miRNA can lead to a variety of diseases;the production of disease is associated with the dysregulation of multiple miRNAs.Therefore,accurate identification of miRNAs regulatory functions can be helpful to understand the mechanism of diseases at the molecular level.It is also of great significance for the biomarker screening,diagnosis,treatment of complex diseases.The number of miRNAs is large,but the proportion of miRNAs with specific regulatory functions is low.The traditional biological experiment method is used to identify the regulatory function between miRNAs and complex diseases,which has the problems of long time and high cost.Therefore,it is an important part of miRNAs research work to study functional similarity between miRNAs by computational methods,identify miRNAs with similar functions,and speculate on the regulatory relationship between miRNAs and complex diseases.In more detail,we propose a novel method to calculate the function between miRNAs,with the goal to address the problems of the mainstream miRNA functional similarity calculation.By exploring the relationship between miRNAs and diseases,we use the regulation information of miRNAs on disease into the miRNAs.Specifically,our method is based on weighted disease semantic information,which inherits the advantages of miRNA functional similarity method based on disease-related information,and further analyzes and utilizes the directional information of miRNAs on disease regulation,in order to calculate the functional similarity of miRNAs more accurately and reasonably.Compared with other three baselines for calculating the functional similarity of miRNAs using disease associations on public datasets(such as HMDD and TAM),our experimental results demonstrate that our method has shown superior performance based on the calculation results.To facilitate the researchers to conduct an in-depth investigation on the function of miRNAs in diseases,the proposed method is implemented as an online service platform for the calculation and analysis of functional similarity of miRNAs.The platform not only provides functions for calculating the functional similarity between miRNAs but also enables visualization of the results.Moreover,it also provides the enrichment analysis function of miRNA sets and prediction of novel miRNA-disease associations based on the functional similarities between miRNAs.The analysis results can also be downloaded for further investigation.In brief,this platform would serve as a convenient tool for miRNA-related researchers.