WEDGE: imputation of gene expression values from single-cell RNA-seq datasets using biased matrix decomposition

Abstract

20The low capture rate of expressed RNAs from single-cell sequencing technology is 21 one of the major obstacles to downstream functional genomics analyses. Recently, a 22 number of recovery methods have emerged to impute single-cell transcriptome profiles, 23 however, restoring missing values in very sparse expression matrices remains a 24 substantial challenge. Here, we propose a new algorithm, WEDGE (WEighted 25Decomposition of Gene Expression), which imputes expression matrix by using a low-26 rank matrix decomposition method. WEDGE successfully restored expression 27 matrices, reproduced the cell-wise and gene-wise correlations, and improved the 28 clustering of cells, performing impressively for applications with multiple cell type 29 datasets with high dropout rates. Overall, this study demonstrates a potent approach 30 for recovering sparse expression matrix data, and our WEDGE algorithm should help 31 many researchers to more profitably explore the biological meanings embedded in 32 their scRNA-seq datasets. 33 34

Single-cell data clustering based on sparse optimization and low-rank matrix factorization

“…For the rank r of gene expression matrix, we use a heuristic algorithm to automatically estimate in scSO. The heuristic algorithm is based on the observation that the ratios of singular values ( ), has a large jump at , and is small ( Hu et al 2020 ). For the covariance structure for GMM, since the features obtained by SNMF are approximately linearly independent, we set the covariance matrix structure to be diagonal for the 12 benchmark datasets.…”

“…First, to reduce the number of free parameters in scSO, we can try to use a hierarchical version of the Expectation–Maximization algorithm to automatically estimate the number of cell clusters in the future version of scSO. Second, recent studies ( Huang et al 2018 ; Hu et al 2020 ) indicated that using an appropriate imputation method for single-cell data can improve the profiling of cell types. As such, we will progress the performance of scSO by enhancing the resistance of scSO to dropout events (due to the low capture and sequencing efficiency of single-cell sequencing technology, most genes are represented by zero values in scRNA-seq data).…”