HDTD

HDTD analyzes high-dimensional transposable data to characterize intra-individual variability in multi-sample datasets such as gene expression measured across multiple tissues within subjects and is implemented within the R/Bioconductor framework.

Key Features:

• Data Structure: Manages matrix-formatted transposable data with phenotypic units indexed by rows (e.g., genes) and subsamples indexed by columns (e.g., tissues).
• Statistical Inference: Provides functions to infer mean relationships between row and column variables and to analyze covariance structures within and between these dimensions.
• Two-way Dependence Modeling: Explicitly addresses the two-way dependence between genes and tissues to avoid biases from methods that ignore transposable structure.
• Hypothesis Testing: Implements hypothesis tests for both mean relationships and covariance structures.
• Platform Implementation: Packaged for use within the R/Bioconductor framework.
• Multi-sample Support: Tailored for datasets with multiple samples per subject, such as multi-tissue gene expression profiles.

Scientific Applications:

• Gene Expression Profiling: Analysis of gene expression across multiple tissues within the same subject to investigate cell type identity and tumor development.
• Intra-subject Variation Analysis: Study of within-subject variability to elucidate physiological processes and disease mechanisms.
• Genotype-Tissue Expression (GTEx) Data Analysis: Application to GTEx datasets for characterizing genetic and phenotypic variability across tissues within individuals.

Methodology:

Performs statistical inference of mean relationships between row and column variables, analyzes covariance structures within and between rows and columns, and conducts hypothesis tests on means and covariance structures to account for two-way dependence in transposable data.