Differential analysis (dana)

Feature-wise stats::lm() or lme4::lmer() models of an omics data matrix. Supports likelihood ratio tests (LRT) and parallel computation.

Usage

dana(
  X,
  sample_data,
  formula_rhs,
  term_LRT = NULL,
  model_control = list(),
  platform = c("ms", "nmr", "ngs"),
  assay = NULL,
  verbose = TRUE
)

Arguments

X: A numeric matrix with samples in rows and features in columns. Sample IDs in row names must match the format from sample_id column in sample_data.
sample_data: A data frame containing sample-level data. Must have a sample_id column matching row names in X and sample_data.
formula_rhs: A one-sided formula (e.g., ~ group + (1|subject)). Must not contain a response variable.
term_LRT: Optional. Character vector of formula terms to test via LRT. Random effects must be written without parentheses (e.g., "1 | group").
model_control: Optional. List of control arguments passed to the model.
platform: Character string indicating the omics platform (e.g., "ms", "nmr", "ngs").
assay: Optional. Character string indicating the name of the platform assay (e.g., "lipidomics").
verbose: Logical. If TRUE, prints progress messages.

Value

An object of class "dana":

X: Matched data matrix.
sdata: Matched sample data.
fit: Data frame of model coefficients and confidence intervals per feature.
lrt: Likelihood ratio test results (if term_LRT is specified).
ranef: Random effects variance components (if using mixed models).
errors: A data frame logging any model fitting errors per feature.

Details

Models are fit independently for each feature using stats::lm() or lmerTest::lmer(), depending on whether dana() detects random effects in formula_rhs. Feature-wise models can be evaluated in parallel using future::plan(), with optional progress updates via progressr::with_progress().

Examples

mock_X <- matrix(
  rnorm(50 * 10) +
    rep(c(rep(0, 25), rep(2, 25)), each = 10) * rep(1:10 %in% 1:3, each = 50),
  nrow = 50
)

rownames(mock_X) <- paste0("sample", 1:50)
colnames(mock_X) <- paste0("feat", 1:10)

sample_data <- data.frame(
  sample_id = rownames(mock_X),
  group = factor(rep(c("A", "B"), each = 25)),
  subject = factor(rep(1:25, each = 2)),
  row.names = rownames(mock_X)
)

# Example with parallel computation setup (not run)
# future::plan(multisession)
# progressr::handlers(global = TRUE)
# progressr::with_progress({
  result <- dana(X = mock_X,
                 sample_data = sample_data,
                 formula_rhs = ~ group + (1 | subject),
                 term_LRT = c("group", "1 | subject"), # Multiple terms allowed
                 platform = "ms",
                 assay = "lipidomics",
                 verbose = FALSE
                 )
#> Computing profile confidence intervals ...
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> boundary (singular) fit: see help('isSingular')
#> Computing profile confidence intervals ...
#> boundary (singular) fit: see help('isSingular')
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> Computing profile confidence intervals ...
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> Computing profile confidence intervals ...
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> Computing profile confidence intervals ...
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> boundary (singular) fit: see help('isSingular')
#> Computing profile confidence intervals ...
#> boundary (singular) fit: see help('isSingular')
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> boundary (singular) fit: see help('isSingular')
#> Computing profile confidence intervals ...
#> boundary (singular) fit: see help('isSingular')
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> Computing profile confidence intervals ...
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> boundary (singular) fit: see help('isSingular')
#> Computing profile confidence intervals ...
#> boundary (singular) fit: see help('isSingular')
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
#> Computing profile confidence intervals ...
#> refitting model(s) with ML (instead of REML)
#> refitting model(s) with ML (instead of REML)
# })

# Modify `dana` object at once with pipes (not run)
# dana_obj <- dana_obj |> adjust_pval() |> add_feat_name() |> ready_plots()

Usage

Arguments

Value

Details

See also

Examples