This repository contains code to explore some of the consequences of using MAGIC in scRNA-seq data.
The report subdirectory contains LaTeX files for the report, which you can compile yourself if you know how.
The scripts subdirectory contains R code to reproduce the analyses described in the report:
nostruct_sim.R, to examine the effect of using MAGIC in a simulation without any structure.structloss_sim.R, involving a simulation containing primary and secondary separations.fakede_sim.R, involving DE analysis between two well-separated subpopulations.run_ercc.R, to examine the effect of using MAGIC on the ERCC data set from 10X Genomics.run_293t.R, to examine the effect of using MAGIC on the 293T data set from 10X Genomics.run_tb.R, to test behaviour when combining B and T cells.