set_MAST {benchdamic}R Documentation

set_MAST

Description

Set the parameters for MAST differential abundance detection method.

Usage

set_MAST(
  pseudo_count = FALSE,
  rescale = c("median", "default"),
  design = NULL,
  coefficient = NULL,
  norm = "TSS",
  expand = TRUE
)

Arguments

pseudo_count

add 1 to all counts if TRUE (default pseudo_count = FALSE).

rescale

Rescale count data, per million if 'default', or per median library size if 'median' ('median' is suggested for metagenomics data).

design

The model for the count distribution. Can be the variable name, or a character similar to "~ 1 + group", or a formula, or a 'model.matrix' object.

coefficient

The coefficient of interest as a single word formed by the variable name and the non reference level. (e.g.: 'ConditionDisease' if the reference level for the variable 'Condition' is 'control').

norm

name of the normalization method used to compute the normalization factors to use in the differential abundance analysis. If norm is equal to "TMM", "TMMwsp", "RLE", "upperquartile", "posupperquartile", "CSSmedian", "CSSdefault", "TSS" the scaling factors are automatically transformed into normalization factors.

expand

logical, if TRUE create all combinations of input parameters (default expand = TRUE)

Value

A named list containing the set of parameters for DA_MAST method.

See Also

DA_MAST

Examples

# Set some basic combinations of parameters for MAST
base_MAST <- set_MAST(design = ~ group, coefficient = "groupB")
# Set a specific set of normalization for MAST (even of other packages!)
setNorm_MAST <- set_MAST(design = ~ group, coefficient = "groupB",
    norm = c("TSS", "poscounts", "TMM"))
# Set many possible combinations of parameters for MAST
all_MAST <- set_MAST(pseudo_count = c(TRUE, FALSE), rescale = c("median",
    "default"), design = ~ group, coefficient = "groupB", norm = c("TSS",
    "poscounts"))

[Package benchdamic version 1.0.0 Index]