PT  - JOURNAL ARTICLE
AU  - Luke Zappia
AU  - Belinda Phipson
AU  - Alicia Oshlack
TI  - Splatter: simulation of single-cell RNA sequencing data
AID  - 10.1101/133173
DP  - 2017 Jan 01
TA  - bioRxiv
PG  - 133173
4099  - http://biorxiv.org/content/early/2017/05/02/133173.short
4100  - http://biorxiv.org/content/early/2017/05/02/133173.full
AB  - As single-cell RNA sequencing (scRNA-seq) technologies have rapidly developed so have methods of analysis. Many of these methods have been tested and developed using simulated datasets. While this is a valid and useful approach many currently published simulations are problematic because they are not well documented, code may not be available for reproducing the simulation or their similarity to real data is not demonstrated.Here we present the Splatter package for simple simulation of single-cell RNA-seq data. Splatter is a Bioconductor R package that provides a consistent, easy to use and well-documented interface for multiple scRNA-seq simulation methods. The Splatter package makes it easy to compare simulated datasets with real data to produce a realistic simulation that can be used to evaluate analysis methods. In addition we develop our own simulation, Splat, based on a gamma-poisson distribution. Splat incorporates a number of key features including high-expression outlier genes, defined library sizes, a mean-variance trend and expression-based dropout. Furthermore, Splat can simulate single populations of cells, populations with multiple cell types or differentiation paths.