netSmooth is an R package for network smoothing of single cell RNA sequencing data. Using gene interaction networks such as protein- protein interactions as priors for gene co-expression, netsmooth improves cell type identification from noisy, sparse scRNA-seq data. The smoothing method is suitable for other gene-based omics data sets such as proteomics, copy-number variation, etc.

The algorithm uses a network-diffusion based approach which takes in a network (such as PPI network) and gene-expression matrix. The gene expression values in the matrix are smoothed using the interaction information in the network. The network-smoothing parameter is optimized using a robust clustering approach.

For a detailed exposition, check out our paper on F1000Research.

netSmooth is available via Bioconductor:

if (!requireNamespace("BiocManager", quietly=TRUE))
	install.packages("BiocManager")
BiocManager::install("netSmooth")

Alternatively, using devtools:

library(devtools)
install_github("BIMSBbioinfo/netSmooth")

For detailed usage information see the vignette. In addition, the R package has full function documentation with examples.

Please cite the netSmooth paper:

Ronen J and Akalin A. netSmooth: Network-smoothing based imputation for single cell RNA-seq [version 2; referees: 2 approved]. F1000Research 2018, 7:8 (doi: 10.12688/f1000research.13511.2)

netSmooth is available under a GPLv3 license.

Fork and send a pull request. Or just e-mail us.

@jonathanronen, BIMSBbioinfo, 2017