Takes a pre-computed PCA (typically calculated on a subset of genes) and projects this onto the entire dataset (all genes). Note that the cell loadings (PCA rotation matrices) remains unchanged, but now there are gene scores for all genes.
ProjectPCA(object, do.print = TRUE, pcs.print = 5, pcs.store = 30,
genes.print = 30, replace.pc = FALSE, do.center = FALSE)Seurat object
Print top genes associated with the projected PCs
Number of PCs to print genes for
Number of PCs to store (default is 30)
Number of genes with highest/lowest loadings to print for each PC
Replace the existing PCA (overwite object@pca.x), not done by default.
Center the dataset prior to projection (should be set to TRUE)
Returns Seurat object with the projected PCA values in object@pca.x.full