herit_vc() and herit_batch() now return var_covariates: the proportion
of phenotypic variance explained by fixed-effect covariates (R² on the
INT-transformed phenotype). This corresponds to the "variance explained"
column reported in Leocadio-Miguel et al. (2025, J Sleep Res). Returns
NA for unadjusted models. sigma2_a and sigma2_e are now documented
using SOLAR's sigma²_g / sigma²_e notation for clarity.
build_grm() — build an additive genetic relationship matrix from a
pedigree data frame via kinship2::kinship(). Supports custom column names,
graceful handling of missing parents, and informative errors for common
mistakes.
herit_vc() — profile-likelihood variance-components heritability estimator
for a single quantitative trait. Features: inverse-normal transformation,
one-sided LRT with chi-squared(1) boundary correction, profile-likelihood
95% CIs, and zero-variance covariate detection.
herit_batch() — iterate herit_vc() over many traits x covariate model
combinations; returns a tidy data frame. Includes a cli progress bar.
int_transform() — exported rank-based inverse-normal transformation
(Blom-style).
plot_forest() — ggplot2 forest plot method for herit_batch() output,
with optional model filtering and significance shading.