Generate HTML report with p-value histograms of all the features

Generates an HTML report based on the results of a limma analysis with splines. The report includes various plots and sections summarizing the analysis results for time effects, average differences between conditions, and interaction effects between condition and time.

Usage

create_limma_report(
  splineomics,
  adj_pthresh = 0.05,
  report_dir = here::here(),
  verbose = TRUE
)

Arguments

splineomics

SplineOmics: An S3 object of class SplineOmics that contains all the necessary data and parameters for the analysis, including:

• limma_splines_result: list A list containing top tables from differential expression analysis for the three different limma results.

• meta: data.frame A data frame with sample metadata. Must contain a column "Time".

• condition: character(1) A character string specifying the column name in the metadata (meta) that defines groups for analysis. This column contains levels such as "exponential" and "stationary" for phases, or "drug" and "no_drug" for treatments.

• annotation: data.frame A data frame containing feature information, such as gene and protein names, associated with the expression data.

• report_info: list A list containing metadata about the analysis for reporting purposes.

adj_pthresh

numeric(1): A numeric value specifying the adjusted p-value threshold for significance. Default is 0.05. Must be > 0 and < 1.

report_dir

character(1): A string specifying the directory where the report should be saved. Default is the current working directory.

verbose

logical(1): Boolean flag controlling the display of messages.

Value

A list of plots included in the generated HTML report.

Examples

set.seed(1)

# --- Toy data: 4 features x 6 samples ---
toy_data <- matrix(
    rnorm(4 * 6),
    nrow = 4, ncol = 6,
    dimnames = list(paste0("feat", 1:4), paste0("S", 1:6))
)

# --- Metadata with required columns (Time, Condition) ---
toy_meta <- data.frame(
    SampleID = colnames(toy_data),
    Time = c(0, 0, 1, 1, 2, 2),
    Condition = factor(c("Ctrl", "Ctrl", "Ctrl", "Trt", "Trt", "Trt"),
        levels = c("Ctrl", "Trt")
    ),
    row.names = colnames(toy_data),
    check.names = FALSE
)

# --- Minimal annotation (feature-level) ---
toy_anno <- data.frame(
    feature_id = rownames(toy_data),
    gene = paste0("G", 1:4),
    row.names = rownames(toy_data),
    check.names = FALSE
)

# --- Helper to fabricate limma-like topTables ---
make_tt <- function(n = 4) {
    p <- runif(n)
    ap <- p.adjust(p, method = "BH")
    data.frame(
        logFC = rnorm(n),
        AveExpr = rnorm(n, 5),
        t = rnorm(n),
        P.Value = p,
        adj.P.Val = ap,
        B = rnorm(n),
        row.names = paste0("feat", 1:n),
        check.names = FALSE
    )
}

# Structure expected by create_limma_report():
toy_limma_res <- list(
    time_effect = list(
        "Time" = make_tt()
    ),
    avrg_diff_conditions = make_tt(),
    interaction_condition_time = make_tt()
)

# --- Build SplineOmics object (note: condition is the COLUMN NAME) ---
so <- create_splineomics(
    data = toy_data,
    meta = toy_meta,
    condition = "Condition",
    annotation = toy_anno,
    design = "1 ~ Condition + Time",
    spline_params = list(spline_type = c("n", "n"), dof = c(2L, 2L)),
    report_info = list(
        omics_data_type       = "RNA-seq (toy)",
        data_description      = "Simulated expression matrix (4x6)",
        data_collection_date  = "2025-10-07",
        analyst_name          = "Analyst A",
        contact_info          = "analyst@example.org",
        project_name          = "SplineOmics Demo"
    )
)

# Attach limma results to the object
so <- update_splineomics(so, limma_splines_result = toy_limma_res)

# --- Generate the HTML report into a temporary directory ---
out_plots <- create_limma_report(
    splineomics = so,
    adj_pthresh = 0.05,
    report_dir  = tempdir(),
    verbose     = FALSE
)