基于loess回归加权的单细胞RNA-seq数据预处理算法  

Single cell RNA-seq data preprocessing algorithm based on LOESS regression weighting

作　　者：高美加 GAO Meijia(College of Computer Science and Technology,Harbin Institute of Technology,Harbin 150040,China)

机构地区：[1]哈尔滨工业大学计算机科学与技术学院,哈尔滨150040

出　　处：《智能计算机与应用》2020年第5期93-97,共5页Intelligent Computer and Applications

摘　　要：单细胞RNA测序数据提供了研究细胞异质性和在生物条件下差异表达基因的机会,其中一些在细胞中表达量有显著变化的高变异基因对单细胞测序数据的下游分析有着关键的作用。本文提出一种基于LOESS回归加权的单细胞RNA-Seq数据预处理算法,处理基因在细胞中的表达量数据,使高变化基因在分析过程中作用加强,达到基因软筛选与数据降噪的目的。进一步,选取6组单细胞RNA-seq数据对算法进行测试,首先对生成的基因表达矩阵进行预处理,然后分析预处理对后续分析(可视化、聚类、差异表达分析)的影响,实验结果表明该算法有效提升了下游分析准确度,显示出良好应用价值。Single-cell RNA-seq data provides us with the opportunity to study cell heterogeneity and differentially expressed genes under biological conditions.Some highly variable genes play a key role in the downstream analysis of single-cell sequencing data.This paper proposes a single-cell RNA-Seq data preprocessing algorithm based on LOESS regression weighting to process gene expression data in cells,so that high-variation genes are strengthened in the entire analysis process to achieve gene soft screening and data noise reduction.Further,I selected 6 single-cell RNA-seq datasets to test the algorithm,first preprocessed the gene expression matrix generated,and then analyzed the impact of pretreatment on subsequent analysis(visualization,clustering,differential expression analysis),experimental results shows that the algorithm effectively improves the accuracy of downstream analysis and shows good application value.

关 键 词：单细胞 RNA测序 数据预处理 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]