The preprocessing pipeline for scRNA-seq has four modules consisting of parameterized Rmarkdown scripts:
- QC+filtering:
prepro_01_qc.Rmd- takes velocyto.loom file and creates seu_raw object (assumes path: "_data/_patients/velocyto/p007n.loom")
- needs filtering parameters for qc (e.g. mito reads)
- Subsampling:
prepro_02_sub.Rmd- takes a Seurat object as input file and outputs a Seurat object as output with N cells
- Anchoring:
prepro_03_anchoring.Rmd- takes multiple qc-filtered Seurat objects and merges (and anchors) them into one merged (and one anchored) Seurat object
- Annotation:
prepro_04_ann.Rmd- takes a seurat object and annotates it with preliminary cell type information and gene expression signatures
Wrapper functions for module calls are described in prepro_00.R. All calls are listed in prepro.sh. R library dependencies are listed within each module file.
After preprocessing, figures were generated as described in scCRC_paper_figures.Rmd.