DMLShrinkage {sesame} | R Documentation |
The function takes a beta value matrix with probes on the rows and samples on the columns. It also takes a sample information data frame (meta) and formula for testing. The function outputs a list of coefficient tables for each factor tested.
DMLShrinkage( betas, formula, meta = NULL, se.lb = 0.06, balanced = FALSE, cf.test = NULL )
betas |
beta values, matrix or SummarizedExperiment |
formula |
formula |
meta |
data frame for sample information, column names are predictor variables (e.g., sex, age, treatment, tumor/normal etc) and are referenced in formula. Rows are samples. |
se.lb |
lower bound to standard error of slope, lower this to get more difference of small effect size. |
balanced |
whether design is balanced or not. default to FALSE, when unbalanced will use Welch's method to estimate standard error. balance=TRUE is faster. |
cf.test |
factors to test (default to all factors in formula except intercept). Use "all" for all factors. |
cf - a list of coefficient tables for each factor
sesameDataCache("HM450") # in case not done yet data <- sesameDataGet('HM450.76.TCGA.matched') cf_list <- DMLShrinkage(data$betas, ~type, meta=data$sampleInfo)