SCnorm and colData in SCE

Question

[TJ]

Hi, I'm looking to normalize my single cell RNA-seq data using SCnorm (Bacher et al. 2017, Nature Methods). My count matrix is coerced into a SingleCellExperiment (sce) container with metadata (for example: genotype and treatment). The metadata was added through scater packer in R (McCarthy et al. 2017, Bioinformatics) using colData. I have been trying unsuccessfully to use the colData for the Conditions function in SCnorm. For example, I have subset my metadata to pull out wildtype and mutant conditions, but continue to get an error:

Condition_1 <- subset(sce, , Genotype=="wildtype") # 226 columns
Condition_2 <- subset(sce, , Genotype=="mutant") # 1101 columns
Conditions = c(rep(Condition_1, 226), rep(Condition_2, 1101)) 

Error: vector memory exhausted(limit reached?)

Ultimately, I have both wildtype and mutant conditions that were treated with Drug A or Drug B presenting 4 total conditions. For now, I am interested in setting the Conditions to two, however, do not know how to do this. Any input would be greatly appreciated. Thank you.

Answer 1

[TJ]

Just a follow up here: Rhonda Bacher kindly provided the answer to this issue, so I figured I would post it here. The Conditions can be subset to metadata that is part of colData in the sce object as follows: Conditions <- sce@colData$cell.conditions In my case, I had 9 different cell conditions with about 1300 cells, it took about 30-45minutes using a single core.