Feature Selection via Correlation or Differential Expression

Selects informative features using either correlation with a quantitative response or differential expression (limma) for binary/continuous responses.

Usage

feature_select(
  x,
  y,
  method = c("cor", "dif"),
  family = c("spearman", "pearson"),
  cutoff = NULL,
  padjcut = NULL
)

Arguments

x

Numeric matrix. Features (rows) by samples (columns).

y

Numeric or factor. Response vector (quantitative or binary).

method

Character. "cor" (correlation) or "dif" (differential expression). Default c("cor","dif").

family

Character. Correlation method if method = "cor": "spearman" or "pearson".

cutoff

Numeric. Absolute correlation (for cor) or |log2FC| (for dif) threshold.

padjcut

Numeric. Adjusted p-value cutoff.

Value

Character vector of selected feature names.

Examples

imvigor210_eset <- load_data("imvigor210_eset")
mad <- apply(imvigor210_eset, 1, mad)
imvigor210_eset <- imvigor210_eset[mad > 0.5, ]
pd1 <- as.numeric(imvigor210_eset["PDCD1", ])
group <- ifelse(pd1 > mean(pd1), "high", "low")
pd1_cor <- feature_select(
  x = imvigor210_eset, y = pd1, method = "cor",
  family = "pearson", padjcut = 0.05, cutoff = 0.5
)
#> Deteching more than two levels in y, we will treat y as a quantitative varibale
pd1_dif <- feature_select(
  x = imvigor210_eset, y = pd1, method = "dif",
  padjcut = 0.05, cutoff = 2
)
#> Deteching more than two levels in y, we will treat y as a quantitative varibale
#> For quantitative varibale, upper 25% and bottom 25% samples
#>               were treated as upregulated group and downregulated group.
pd1_dif_2 <- feature_select(
  x = imvigor210_eset, y = group,
  method = "dif", padjcut = 0.05, cutoff = 2
)
#> Deteching two levels in y, we will treat y as a binary varibale