netNMF-sc: leveraging gene–gene interactions for imputation and dimensionality reduction in single-cell expression analysis

Abstract

Single-cell RNA-sequencing (scRNA-seq) enables high-throughput measurement of RNA expression in single cells. However, because of technical limitations, scRNA-seq data often contain zero counts for many transcripts in individual cells. These zero counts, or dropout events, complicate the analysis of scRNA-seq data using standard methods developed for bulk RNA-seq data. Current scRNA-seq analysis methods typically overcome dropout by combining information across cells in a lower-dimensional space, leveraging the observation that cells generally occupy a small number of RNA expression states. We introduce netNMF-sc, an algorithm for scRNA-seq analysis that leverages information across both cells and genes. netNMF-sc learns a low-dimensional representation of scRNA-seq transcript counts using network-regularized non-negative matrix factorization. The network regularization takes advantage of prior knowledge of gene-gene interactions, encouraging pairs of genes with known interactions to be nearby each other in the low-dimensional representation. The resulting matrix factorization imputes gene abundance for both zero and nonzero counts and can be used to cluster cells into meaningful subpopulations. We show that netNMF-sc outperforms existing methods at clustering cells and estimating gene-gene covariance using both simulated and real scRNA-seq data, with increasing advantages at higher dropout rates (e.g., >60%). We also show that the results from netNMF-sc are robust to variation in the input network, with more representative networks leading to greater performance gains.

scGNN is a novel graph neural network framework for single-cell RNA-Seq analyses

“…3 ). Median L1 distance, cosine similarity, and root-mean-square-deviation (RMSE) scores between the original data set and the imputed values for these synthetic entries were calculated to compare scGNN with MAGIC 4 , SAUCIE 10 , SAVER 19 , scImpute 33 , scVI 32 , DCA 11 , DeepImpute 34 , scIGANs 35 , and netNMF-sc 36 (see “Methods” section). scGNN achieves the best results in recovering gene expressions in terms of median L1 distance, and RMSE at the 10 and 30% synthetic dropout rate, respectively.…”

“…The process is repeated three times, and the mean and standard deviation were selected as a comparison. The scores are compared between scGNN and nine imputation tools (i.e., MAGIC 4 , SAUCIE 10 , SAVER 19 , scImpute 33 , scVI 32 , DCA 11 , DeepImpute 34 , scIGANs 35 , and netNMF-sc 36 ), using the default parameters.…”

Statistics or biology: the zero-inflation controversy about scRNA-seq data

“…Such excess zeros would bias the estimation of gene expression correlations [42] and hinder the capture of gene expression dynamics [43] from scRNAseq data. In early scRNA-seq data analyses, the high data sparsity provoked the use of zeroinflated models [36,38,44] and the development of imputation methods for reducing zeros [20,42,43,[45][46][47][48][49][50][51][52][53][54][55][56][57][58][59][60]. More recently, however, there were voices against the use of zero-inflated models for scRNA-seq data generated by UMI protocols [61].…”

Robust and accurate deconvolution of tumor populations uncovers evolutionary mechanisms of breast cancer metastasis

“…While much of the motivation for this work is the difficulty of acquiring scRNA for primary tumor samples when examining metastases years later, we do anticipate that this problem will lessen over time. It is thus worth considering for the future whether our methods might be adapted for working on limited and noisy scRNA with matched bulk data ( Elyanow et al , 2020 ).…”