Normalised Chromosome Value (NCV) score
nipter_ncv_score.RdComputes the NCV score for a test sample. The NCV method selects an optimal set of denominator chromosomes that minimises the coefficient of variation (CV) in the control group, then uses the resulting ratio to Z-score the test sample.
Usage
nipter_ncv_score(
sample,
control_group,
chromo_focus,
max_elements = 5L,
exclude_chromosomes = c(13L, 18L, 21L),
include_chromosomes = NULL
)Arguments
- sample
A
NIPTeRSampleobject.- control_group
A
NIPTeRControlGroupobject.- chromo_focus
Integer; the target chromosome (1-22).
- max_elements
Maximum number of denominator chromosomes to try (default 5). The algorithm searches all combinations of 1 to
max_elementsfrom the candidate pool.- exclude_chromosomes
Integer vector of chromosomes to exclude from the denominator pool (default
c(13, 18, 21)— the trisomy chromosomes).- include_chromosomes
Integer vector of chromosomes to force-include in the candidate pool. Default
NULL.
Value
A list of class "NIPTeRNCV" with elements:
- sample_score
The NCV score for the test sample (numeric).
- focus_chromosome
Character; the tested chromosome.
- denominators
Integer vector of selected denominator chromosomes.
- control_statistics
Named numeric vector with
mean,sd, andshapiro_p_value.- control_z_scores
Named numeric vector of control NCV scores.
- best_cv
The minimum CV achieved.
- correction_status
Character vector.
- sample_name
Character.
Details
The candidate denominator pool is: chromosomes 1-12, 14-17, 19-20, 22
(the "control chromosomes") minus exclude_chromosomes minus
chromo_focus, plus any include_chromosomes.
For each combination size from 1 to max_elements, all
choose(n_candidates, size) combinations are evaluated. The ratio
reads[focus] / sum(reads[denominators]) is computed per control
sample, and the combination minimising the coefficient of variation
(sd/mean) is selected.