Application of Sparse Representation in Bioinformatics.

Shuguang Han,Ying Ju,Ning Wang,Furong Tang,Lei Xu,Lei Shi,Yuxin Guo

doi:10.3389/fgene.2021.810875

Abstract

Inspired by L1-norm minimization methods, such as basis pursuit, compressed sensing, and Lasso feature selection, in recent years, sparse representation shows up as a novel and potent data processing method and displays powerful superiority. Researchers have not only extended the sparse representation of a signal to image presentation, but also applied the sparsity of vectors to that of matrices. Moreover, sparse representation has been applied to pattern recognition with good results. Because of its multiple advantages, such as insensitivity to noise, strong robustness, less sensitivity to selected features, and no “overfitting” phenomenon, the application of sparse representation in bioinformatics should be studied further. This article reviews the development of sparse representation, and explains its applications in bioinformatics, namely the use of low-rank representation matrices to identify and study cancer molecules, low-rank sparse representations to analyze and process gene expression profiles, and an introduction to related cancers and gene expression profile database.

Highlights

In recent years, inspired by L1-norm minimization methods, such as basis pursuit (Donoho and Huo, 2001), compressed sensing (Candes et al, 2004; Candes and Tao, 2005; Lustig et al, 2007), and Lasso feature selection (Tibshirani, 1996), sparse representation shows up as a novel and potent data processing method
Gene expression profile data analysis has attracted widespread attention from scholars, and a series of gene expression profile analysis methods have been proposed. Classic methods such as principal component analysis (PCA), LDA, KNN, decision-making tree method, ensemble learning, support vector machine (SVM), extreme learning machine, neural network, sparse representation, and gene bi-clustering method based on qualitative/quantitative measurement have been widely applied to the classification and clustering of gene expression profile data
A dimensionality reduction method based on low-rank graphs is a more effective feature extraction method

Summary

Introduction

In recent years, inspired by L1-norm minimization methods, such as basis pursuit (Donoho and Huo, 2001), compressed sensing (Candes et al, 2004; Candes and Tao, 2005; Lustig et al, 2007), and Lasso feature selection (Tibshirani, 1996), sparse representation shows up as a novel and potent data processing method. The effectiveness of this method has been widely recognized in image processing, and it provides new ideas and directions for establishing accurate models for mining cancer molecular characteristics. A mathematical model based on low-rank representation can be established by combining multiple scales, including molecules, modules, functional networks, and multi-omics molecular features.

Results

Conclusion