Run the NIPTmer regression model on count data — niptmer_score

This ports the legacy SEX Inference model into stateless R utilities.

Usage

niptmer_score_samples(
  count_table,
  control = NULL,
  normalize = FALSE,
  extra_columns = character(),
  boy_threshold = 0.02,
  girl_threshold = 0.01,
  reference_mode = c("fixed", "match_subset", "filtered_controls"),
  reference_top_n = 100,
  reference_exclude_chromosomes = c("13", "18", "21", "X", "Y", "GC"),
  reference_diagnose_chromosomes = as.character(seq_len(22L)),
  reference_z_threshold = 3,
  reference_max_outlier_chromosomes = 1
)

Arguments

count_table: A NIPT count table data.frame or file path.
control: Either NULL, a control file path, a character vector of control sample names, or a 1-2 column data.frame (sample, optional sex).
normalize: Whether to use normalized differences (actual - predicted) / predicted.
extra_columns: Optional additional predictor columns.
boy_threshold: Y-coverage threshold above which sample is called boy.
girl_threshold: Y-coverage threshold below which sample is called girl.
reference_mode: Control-group strategy: "fixed" uses one shared control group; "match_subset" picks best-matching controls per sample via Match QC; "filtered_controls" removes outlier controls before scoring.
reference_top_n: Number of controls for sample-specific matching when reference_mode = "match_subset".
reference_exclude_chromosomes: Chromosomes excluded from Match QC.
reference_diagnose_chromosomes: Chromosomes used for control diagnostics.
reference_z_threshold: Absolute z-score threshold for control diagnostics.
reference_max_outlier_chromosomes: Maximum allowed outlier chromosomes in control diagnostics.

Value

A data.frame with autosomal/GC z-scores, sex-model z-scores, Mahalanobis distance, and model gender.

Examples

counts <- system.file("extdata", "niptmer_counts_example.tsv", package = "RGenomeTester4")
controls <- system.file("extdata", "niptmer_controls_example.tsv", package = "RGenomeTester4")
scores <- niptmer_score_samples(counts, control = controls, normalize = TRUE)
niptmer_interpret(scores)
#>   SampleName AA SA Distance Gender
#> 1       2NV9    XY     Good    boy