Preprocessing

Seurat pipeline

Background and sources

Single-cell RNA-seq methods develop rapidly and no standardized workflow exists. This analysis is based on current best practices as outlined by Luecken and Theis (June 2019). Major analysis platforms are Seurat and Scater (written in R) and scanpy (written in Python). A web-based course and overview of scRNA-seq methods, including Seurat, is maintained by the Hemberg lab.

We use the Seurat package maintained by the Satija lab and available through CRAN. The version used here is Seurat 3.0 (Released April 16, 2019). See their guided tutorials and the review above for detailed explanations. Packages and algorithms used are linked to published articles or repositories for further information.

Loading packages and setting main parameters

First we load packages required for the main analysis, plotting and for later pseudotime ordering using multiple different algorithms.

At this step we also decide if cell cycle affects should be regressed using the filter_cell_cycle parameters.

The user must supply a working directory to use for importing data and generating outputs.