GPU (Graphics Processing Unit)-based rapid optimization operation gene co-expression network method

Abstract

The invention discloses a GPU (Graphics Processing Unit)-based rapid optimization operation gene co-expression network method, which belongs to the field of internet and comprises the following steps: 1, establishing a data reading module, a correlation matrix coefficient iteration module and a grade numerical value iteration module, providing computing power by using a GPU, and constructing the gene co-expression network; and 2, reading related data of the input file by using a data reading module. An iteration idea is introduced, and a new value is recursively calculated by using an old value, so that the new value can adapt to gene co-expression data under different conditions, and then a weighted association network is constructed and a result with more biological significance is obtained; and the GPU has multiple cores and is applied to gene co-expression network analysis to process a gene co-expression matrix, so that the analysis period can be greatly shortened, and the analysis waiting time can be shortened. In this way, needed hardware and software environments are simple, and installation is convenient.

Description

technical field [0001] The invention relates to the technical field of the Internet, in particular to a method for quickly optimizing and computing a gene co-expression network based on GPU. Background technique [0002] Since gene co-expression data usually measure the expression values ​​of tens of thousands of genes in dozens of samples, for such a huge matrix, the calculation of gene data processing is extremely large and the efficiency is not high, so many bioinformatics tools have been developed Using network biology methods to analyze and mine high-throughput gene expression data has become an important research direction of bioinformatics. The gene co-expression matrix is ​​a multidimensional matrix, and the computational optimization of multidimensional matrices and complex networks has always been a hot topic for researchers at home and abroad. [0003] Rudolph et al. conducted research on the directed WS small-world network model, and proposed a new analytical fr...

AI Technical Summary

Problems solved by technology

However, the efficiency of dimensionality reduction is low, and the difficulty of dimensionality reduction makes these methods unsatisfactory.

Images