ExperimentSubset

ExperimentSubset enables efficient, provenance-preserving subsetting and management of Bioconductor Experiment objects (e.g., SummarizedExperiment, SingleCellExperiment) that store matrix-like assays with associated row and column metadata from technologies such as single-cell RNA sequencing.

Key Features:

• Efficient subsetting: Performs flexible row- and column-wise subsetting of assay matrices without creating duplicate assay objects.
• Data provenance: Maintains an explicit link between subsetted assays and their parent assay to track the origin of subsets.
• Memory efficiency: Avoids duplication of original assays to reduce memory usage for large-scale genomic datasets such as scRNA-seq.
• Experiment class compatibility: Provides an interface consistent with Bioconductor Experiment classes to preserve compatibility with Experiment-based analysis frameworks.

Scientific Applications:

• High-throughput genomic data management: Supports handling and organization of matrix-like assay data with associated row and column metadata in genomic studies.
• Preprocessing and quality control: Enables exclusion of poor-quality samples and selection of highly variable features during preprocessing steps.
• Single-cell RNA-seq workflows: Facilitates memory-efficient subsetting and feature/sample selection in large single-cell RNA sequencing datasets.
• Provenance-aware downstream analysis: Allows tracking of subset origins across successive analysis stages to support reproducibility.

Methodology:

Implemented as an R package that defines specialized classes to store subsetted assays within the original Experiment object framework (e.g., SummarizedExperiment, SingleCellExperiment), maintaining links to parent assays for provenance tracking and avoiding data duplication to optimize memory usage.