Create PVC HTML report from statistical PVC results

Generates per-feature PVC plots and writes an HTML report to disk from the statistical output of find_pvc(). This function does not recompute any statistics; it consumes pvc_results (including its attributes) and the original input data stored in splineomics.

Usage

create_pvc_report(
  splineomics,
  pvc_results,
  plot_info = list(y_axis_label = "Value", time_unit = "min", treatment_labels = NA,
    treatment_timepoints = NA),
  verbose = FALSE,
  report_dir = tempdir()
)

Arguments

splineomics

list: Preprocessed time-series input. Must include at least data, meta, annotation, condition, meta_batch_column, meta_batch2_column (optional), report_info, and feature_name_columns.

pvc_results

list: Output from find_pvc(), a named list by condition level. Each level must contain alpha and pvc_adj_pvals. Attributes padjust_method and support are used for report header settings.

plot_info

list: Plot annotation options passed to plot_pvc(). See find_pvc() for the expected structure. Treatment annotations are applied per condition level if treatment_labels and treatment_timepoints are provided.

verbose

Boolean flag indicating if messages should be shown.

report_dir

character(1): Output directory for the HTML report and any associated files.

Value

The input pvc_results with an additional plots element added under each condition level. The report is written to report_dir.

Details

For each condition level, the function subsets splineomics$data and splineomics$meta, generates per-feature plots via plot_pvc(), then calls generate_report_html() with the plots, metadata, and report header settings derived from pvc_results and splineomics.

Examples

## Runnable example using packaged extdata (small subset for speed)
data_in <- readRDS(
  xzfile(
    system.file(
      "extdata",
      "phosphoproteomics_data.rds.xz",
      package = "SplineOmics"
    )
  )
)

meta <- read.csv(
  system.file(
    "extdata",
    "phosphoproteomics_meta.csv",
    package = "SplineOmics"
  ),
  stringsAsFactors = FALSE
)

first_na_col <- which(is.na(data_in[1, ]))[1]
annotation_all <- data_in[, (first_na_col + 1):ncol(data_in), drop = FALSE]

## Extract a small feature subset to keep examples fast
data_mat <- SplineOmics::extract_data(
  data = data_in,
  feature_name_columns = c("T..Gene"),
  use_row_index = TRUE,
  top_row = 1,
  bottom_row = 50,
  right_col = 36,
  left_col = 1
)

## IMPORTANT: subset annotation to match the selected features
keep_ids <- rownames(data_mat)
annotation <- annotation_all[
    annotation_all$T..Gene %in% keep_ids, , drop = FALSE
    ]
annotation <- annotation[
    match(keep_ids, annotation$T..Gene), , drop = FALSE
    ]

splineomics <- SplineOmics::create_splineomics(
  data = data_mat,
  meta = meta,
  annotation = annotation,
  condition = "Phase",
  meta_batch_column = "Reactor"
)

pvc_results <- SplineOmics::find_pvc(
  splineomics = splineomics,
  alphas = 0.025,
  padjust_method = "BH",
  support = 1
)
#> design matrix of interest not specified. Assuming a one-group experiment.
#> Warning: Partial NA coefficients for 2 probe(s)
#> Warning: Partial NA coefficients for 12 probe(s)
#> design matrix of interest not specified. Assuming a one-group experiment.
#> Warning: Partial NA coefficients for 2 probe(s)
#> Warning: Partial NA coefficients for 9 probe(s)

report_dir <- file.path(tempdir(), "pvc_report_example")
dir.create(report_dir, showWarnings = FALSE, recursive = TRUE)

out <- SplineOmics::create_pvc_report(
  splineomics = splineomics,
  pvc_results = pvc_results,
  verbose = FALSE,
  report_dir = report_dir
)

names(out)
#> [1] "Exponential" "Stationary"