DA_ANCOM: DA_ANCOM

In mcalgaro93/benchdamic: Benchmark of differential abundance methods on microbiome data

DA_ANCOM

Description

Fast run for ANCOM and ANCOM-BC2 differential abundance detection methods.

Usage

DA_ANCOM(
  object,
  assay_name = "counts",
  pseudo_count = FALSE,
  fix_formula = NULL,
  adj_formula = NULL,
  rand_formula = NULL,
  lme_control = lme4::lmerControl(),
  contrast = NULL,
  p_adj_method = "BH",
  struc_zero = FALSE,
  BC = TRUE,
  n_cl = 1,
  verbose = TRUE
)

Arguments

object	a phyloseq or TreeSummarizedExperiment object.
assay_name	the name of the assay to extract from the TreeSummarizedExperiment object (default assayName = "counts"). Not used if the input object is a phyloseq.
pseudo_count	add 1 to all counts if TRUE (default pseudo_count = FALSE).
fix_formula	Used when BC = TRUE (ANCOM-BC2). The character string expresses how the microbial absolute abundances for each taxon depend on the fixed effects in metadata.
adj_formula	Used when BC = FALSE (ANCOM). The character string represents the formula for covariate adjustment. Default is NULL.
rand_formula	Optionally used when BC = TRUE or BC = FALSE. The character string expresses how the microbial absolute abundances for each taxon depend on the random effects in metadata. ANCOMB and ANCOM-BC2 follows the lmerTest package in formulating the random effects. See ?lmerTest::lmer for more details. Default is rand_formula = NULL.
lme_control	a list of control parameters for mixed model fitting. See ?lme4::lmerControl for details.
contrast	character vector with exactly, three elements: a string indicating the name of factor whose levels are the conditions to be compared, the name of the level of interest, and the name of the other level.
p_adj_method	character. method to adjust p-values. Default is "holm". Options include "holm", "hochberg", "hommel", "bonferroni", "BH", "BY", "fdr", "none". See ?stats::p.adjust for more details.
struc_zero	logical. Whether to detect structural zeros based on group. Default is FALSE. See Details for a more comprehensive discussion on structural zeros.
BC	boolean for ANCOM method to use. If TRUE the bias correction (ANCOM-BC2) is computed (default BC = TRUE). When BC = FALSE computational time may increase and p-values are not computed.
n_cl	numeric. The number of nodes to be forked. For details, see ?parallel::makeCluster. Default is 1 (no parallel computing).
verbose	an optional logical value. If TRUE, information about the steps of the algorithm is printed. Default verbose = TRUE.

Value

A list object containing the matrix of p-values 'pValMat', a matrix of summary statistics for each tag 'statInfo', and a suggested 'name' of the final object considering the parameters passed to the function. ANCOM (BC = FALSE) does not produce p-values but W statistics. Hence, 'pValMat' matrix is filled with 1 - W / (nfeatures - 1) values which are not p-values. To find DA features a threshold on this statistic can be used (liberal < 0.4, < 0.3, < 0.2, < 0.1 conservative).

See Also

ancombc for analysis of microbiome compositions with bias correction or without it ancom.

Examples

set.seed(1)
# Create a very simple phyloseq object
counts <- matrix(rnbinom(n = 60, size = 3, prob = 0.5), nrow = 10, ncol = 6)
metadata <- data.frame("Sample" = c("S1", "S2", "S3", "S4", "S5", "S6"),
                       "group" = as.factor(c("A", "A", "A", "B", "B", "B")))
ps <- phyloseq::phyloseq(phyloseq::otu_table(counts, taxa_are_rows = TRUE),
                         phyloseq::sample_data(metadata))
# Differential abundance
DA_ANCOM(object = ps, pseudo_count = FALSE, fix_formula = "group", contrast =
   c("group", "B", "A"), verbose = FALSE)

mcalgaro93/benchdamic documentation built on Nov. 28, 2022, 6:55 p.m.