MiRNA-disease incidence relation prediction method based on multi-source data fusion and matrix completion

Abstract

The invention discloses a miRNA-disease association relationship prediction method based on multi-source data fusion and matrix completion. The method comprises the following steps: firstly, obtaining related data of miRNA and diseases from a plurality of known biological databases; the method comprises the following steps: constructing a miRNA self-adjacency matrix, a disease self-adjacency matrix and a miRNA-disease adjacency matrix by using different data source information; analyzing a potential association relationship between the miRNA self-adjacency matrix and the disease self-adjacency matrix by using a matrix completion algorithm; calculating a kernel similarity matrix of a miRNA Gaussian interaction spectrum and a kernel similarity matrix of a disease Gaussian interaction spectrum through the miRNA-disease adjacency matrix; a plurality of processed adjacent matrixes are combined into a miRNA-disease association network; and obtaining a final prediction result on the associated network by using a label propagation algorithm. The method is high in prediction precision and low in cost, aims to excavate the incidence relation between miRNA and diseases, and can provide support for exploring pathogenesis of complex diseases.

Description

technical field [0001] The invention relates to the field of bioinformatics, in particular to a miRNA-disease correlation prediction method based on multi-source data fusion and matrix completion. Background technique [0002] miRNA (microRNA) is an endogenous RNA, which is a small RNA composed of 22-24 nucleotides. miRNAs complete gene regulation functions by cutting off the mRNA of the target gene to complement the target gene or by inhibiting the translation of the target gene. In recent years, more and more studies have shown that the expression of miRNAs may have potential relationships with many complex diseases. The first confirmed miRNAs were lin-4 and let-7 from C. elegans, and researchers have been advancing the exploration of miRNAs since then. Nowadays, there is a lot of evidence that miRNAs are involved in a series of human life activities. The early development, proliferation, differentiation, and death of human cells are closely related to miRNAs. [0003] ...

AI Technical Summary

Problems solved by technology

However, these methods use a relatively sparse relationship matrix for calculation, which affects the final prediction results

Images