anpan_repeated_measures: Use repeated measures to refine the gene model

In biobakery/anpan: Quantifying Microbial Strain-Host Associations

Use repeated measures to refine the gene model

Description

Use repeated measures to refine the gene model

Usage

anpan_repeated_measures(
  subject_sample_map,
  bug_dir,
  meta_file,
  out_dir,
  genomes_dir = NULL,
  model_type = "fastglm",
  covariates = c("age", "gender"),
  outcome = "crc",
  omit_na = FALSE,
  filtering_method = "kmeans",
  minmax_thresh = NULL,
  discard_poorly_covered_samples = TRUE,
  skip_large = TRUE,
  save_fit = TRUE,
  annotation_file = NULL,
  save_filter_stats = TRUE,
  verbose = TRUE,
  plot_result = TRUE,
  plot_ext = "pdf",
  q_threshold = 0.1,
  n_top = 50,
  width = 10,
  height = 8,
  ...
)

Arguments

subject_sample_map	a data frame between the mapping between subject_id and sample_id
bug_dir	a directory of gene family files
meta_file	path to a metadata tsv
out_dir	path to the desired output directory
genomes_dir	an optional directory of genome files
model_type	either "horseshoe" or "fastglm"
covariates	character vector of covariates to include in the model
outcome	character string of the outcome variable
omit_na	logical indicating whether to omit incomplete cases of the metadata
filtering_method	method to use for filtering samples. Either "kmeans" or "none"
minmax_thresh	genes must have at least this many (or N - this many) non-zero observations or else be discarded. NULL defaults to floor(0.005*nrow(metadata)).
discard_poorly_covered_samples	logical indicating whether to discard samples where the genes of a bug are poorly covered
skip_large	logical indicating whether to skip bugs with over 5k genes. Only used when model_type = "horseshoe".
save_fit	logical indicating whether to save horseshoe fit objects. Only used when model_type = "horseshoe".
annotation_file	a path to a file giving annotations for each gene
save_filter_stats	logical indicating whether to save filter statistics
plot_result	logical indicating whether or not to plot the results
plot_ext	extension to use for plots
q_threshold	FDR threshold to use for inclusion in the plot.
n_top	number of top elements to show from the results
width	width of saved plot in inches
height	height of saved plot in inches
...	arguments to pass to [cmdstanr::sample()] if applicable

Details

This function performs the standard anpan filtering on all samples, then uses the subject-sample map to compute the proportion of samples with the bug. This gives a gene _proportion_ matrix (instead of a presence/absence matrix) which is then passed to anpan_batch(filtering_method = "none", discretize_inputs = FALSE). Subjects that do not have the bug present in at least half their samples are dropped.

In cases where subject metadata varies by sample, the mean is taken if the variable is numeric, otherwise it is tabulated and the most frequent category is selected as the subject-level metadata value. This tabulation will respect factor ordering if you'd like to alter the value selected in the event of ties.

biobakery/anpan documentation built on Aug. 14, 2024, 8:19 a.m.