RaggedExperiment

RaggedExperiment represents lossless “ragged” genomic ranges in R/Bioconductor by encoding per-sample or per-cell genomic intervals to enable computation of rectangular summaries for downstream analyses of somatic mutations, copy number variation, methylation, and chromatin accessibility.

Key Features:

• Lossless representation: Retains full input genomic coordinates and associated metadata without compression or approximation.
• Ragged-range storage: Stores irregular genomic intervals that differ across samples or cells in a coherent data structure.
• Reshaping and aggregation: Provides flexible reshaping operations to convert heterogeneous intervals into tabular formats via user-selectable aggregation and summarization strategies.
• Matrix-like summaries: Enables efficient computation of rectangular (matrix-like) summaries required for downstream statistical modeling.
• Interoperability: Operates with Bioconductor’s MultiAssayExperiment class to support multimodal data integration.

Scientific Applications:

• Somatic mutation analysis: Representation and summarization of somatic mutations and mutation burden estimation.
• Copy number variation: CNV profiling and summarization of sample-specific copy number intervals.
• Methylation and chromatin accessibility: Handling and summarization of methylation and chromatin accessibility assay-derived intervals.
• SNP-based modeling and integrative multiomics: Support for SNP-based modeling and integration across somatic mutation, CNV, methylation, and chromatin accessibility data via MultiAssayExperiment.
• TCGA use: Demonstrated application across 33 TCGA tumor cohorts.

Methodology:

Reshaping operations and user-selectable aggregation and summarization strategies convert heterogeneous genomic intervals into tabular, rectangular (matrix-like) summaries while retaining exact coordinates and metadata and interoperating with MultiAssayExperiment.