Changelog

Source: NEWS.md

moat 0.99.0

• Renamed and repositioned the package as moat, the Microbiome Omics Audit Toolkit. MOAT is framed as a pre-analysis audit layer: before crossing into microbiome analysis, check the moat.
• Added moat_audit, summary.moat_audit, and moat_analysis_plan as the package object classes.
• Replaced remaining legacy condition classes and audit-object helpers with moat_* names; no compatibility aliases are provided.
• Added tidy audit extraction helpers and a programmatic risk-threshold reference for reports and methods documentation.

New Package

• Initial development release prepared for future Bioconductor submission.
• Added the moat_audit object returned by moat().
• Added input validation for SummarizedExperiment-like microbiome objects.
• Added toy microbiome data for examples, tests, and vignettes.

Design And Correction Audits

• Added check_design() for metadata associations with the outcome.
• Added check_continuous_design() for continuous covariates.
• Added check_balance() and check_model_matrix() for correction feasibility diagnostics.
• Added check_correction() to summarize whether batch adjustment is statistically identifiable.

Microbiome Batch Audits

• Added transform_biome() and compute_biome_distance() for supported microbiome transformations and distances.
• Added check_permanova() for distance-based PERMANOVA diagnostics.
• Added check_batch() for PERMANOVA, dispersion, and PCoA-axis batch audits across selected distances.
• Added optional PERMANOVA term-order sensitivity diagnostics for outcome-first versus batch-first model formulas.
• Added check_feature_batch() to screen individual microbiome features for batch association and include compact feature-level summaries in batch audits.

Leakage And Risk Scoring

• Added check_repeated_measures() and check_leakage() for subject, batch, and time-aware validation risk.
• Added central risk scoring and summary.moat_audit().
• Added conservative low, moderate, high, critical, and unknown risk handling.

Analysis Planning And Plotting

• Added plan_analysis() for downstream formula, validation, batch strategy, and sensitivity recommendations.
• Added plot_design() to visualize outcome distribution across audited categorical metadata variables.
• Added plot_variance() to visualize PERMANOVA R2 terms and batch dominance.
• Added report() to render offline HTML reports from moat_audit objects.

Documentation And Release Preparation

• Added an introductory vignette covering the full audit workflow.
• Added a vignette showing how pre-analysis audit changes interpretation for batch-dominated beta diversity, metadata-only predictability, and leakage.
• Added a validation-leakage vignette for microbiome machine-learning workflows.
• Updated README with current user-facing APIs.
• Made source-only documentation workflow tests skip cleanly in installed package checks and stopped tracking generated vignette HTML artifacts.
• Added citation metadata placeholder for the development package.