Filter & Normalize

Filter the Data

Visualized QC metrics are then used to filter cells.

Remove cells with unique features (genes) over 2,500 or less than 200, and remove cells that have greater than 5% mitochondrial counts.

Note that subset() arguments define cells you want to keep (e.g., percent.mt < 5).

# Subset cells for further analysis based on QC metrics. 
pbmc <- subset(pbmc, subset = nFeature_RNA > 200 & nFeature_RNA < 2500 & percent.mt < 5)

Normalize the Data

Use global-scaling normalization (LogNormalize).

For each cell, (a) divide feature expression counts by the total counts for that cell, (b) multiply by a scale factor (default is 10,000), and (c) apply a log transformation.

The normalized data is stored in pbmc[["RNA"]]@data.

pbmc <- NormalizeData(pbmc, normalization.method = "LogNormalize", scale.factor = 10000)