OmicsBraid: Covariance-Aware Inference of Cross-Omic Effect Trajectories
A research-oriented statistical framework for comparing standardized
biological effects across matched omics layers. It estimates layer-specific
standardized effects, accounts for cross-omic dependence using matched-subject
bootstrap correlations, tests multivariate omnibus evidence, synthesizes
consensus effects with generalized least squares, quantifies cross-omic
heterogeneity, performs practical-equivalence testing, fits covariance-aware
ordered GLS effect trajectories, classifies hierarchical cross-layer effect
patterns with separate confirmatory and suggestive states, supports analytic
and subject-bootstrap confidence intervals for layer and consensus effects,
supports empirical matched-subject permutation and centered-bootstrap
calibration of omnibus and heterogeneity tests for non-Gaussian settings,
and creates evidence-forest and effect-braid visualizations. The package is
designed for analysis-ready bulk multi-omics data or externally estimated
summary statistics. It does not perform raw sequencing or mass-spectrometry
preprocessing. Methodological components draw on standardized mean-difference
estimation described by Hedges (1981) <doi:10.3102/10769986006002107>,
bootstrap resampling described by Efron (1979)
<doi:10.1214/aos/1176344552>, and two one-sided equivalence testing
described by Schuirmann (1987) <doi:10.1007/BF01068419>.
Documentation:
Downloads:
Linking:
Please use the canonical form
https://CRAN.R-project.org/package=OmicsBraid
to link to this page.