An Open, Extensible, Benchmark Platform To Evaluate Computational Tools for Single-Cell Differential Expression

To generate an open, reproducible and extensible benchmark platform for evaluating computational tools for single-cell differential expression analysis.

The Stephens lab developed an initial implementation of a dynamic benchmark to assess methods for detecting differential expression in single-cell datasets, leading to improvements in their benchmarking software. As part of the benchmarking process they built on methods for simulating useful datasets for comparing differential expression methods, which are implemented in the R package seqgendiff.