Generate a matrix of anova values for associating principal components with categorical covariates.

Usage

anova_pca_metadata(pca_coords, pcameta, fraction_explained, n_components = 10)

Arguments

pca_coords: Data frame of PCA coordinates, with samples by row and components by column.
pcameta: Data frame of sample metadata with sample identifiers by row and variables by column.
fraction_explained: Numeric vector containing the percent contribution to variance of each component
n_components: Number of leading components to test. Clamped to the number actually available in pca_coords if that's fewer.

Value

output A numeric matrix of p values

Examples

mat <- matrix(rnorm(90), nrow = 15, ncol = 6,
  dimnames = list(paste0("gene", 1:15), paste0("s", 1:6)))
pca <- compile_pca_data(mat)
pcameta <- data.frame(
  condition = rep(c("treated", "control"), each = 3),
  batch = rep(c("A", "B", "C"), 2),
  row.names = colnames(mat)
)
anova_pca_metadata(pca$coords, pcameta, pca$percentVar)
#>           PC1 (36.2%) PC2 (25%) PC3 (16.4%) PC4 (13.4%)  PC5 (9%)  PC6 (0%)
#> condition 0.009998015 0.9528623 0.995902831   0.4398678 0.9224774 0.4085805
#> batch     0.976610515 0.6584137 0.002567231   0.9643998 0.1356075 0.1114647