Biobase

Functions that are needed by many other packages or which replace R functions.

Dependencies & Environment

Package-intrinsic requirements from the Bioconductor landing page — reproduce in any R environment.

• Version: 2.72.0 · Bioconductor: 3.23 · R: ≥ 4.6
• Depends: BiocGenerics
• Install: BiocManager::install("Biobase")

When to Use

• Coordinating high-throughput genomic data (e.g., microarray expression values) and phenotype metadata into a single, structured ExpressionSet object.
• Evaluating statistical functions across sample strata defined by covariates using esApply.
• Updating older, serialized instances of Bioconductor objects to their current class representations using updateObject().

When NOT to Use

• For modern single-cell or range-based sequencing data, use SummarizedExperiment or SingleCellExperiment instead because they natively support genomic coordinates and scale better to sparse, multi-assay datasets.
• For purely tabular data manipulation, use dplyr or data.table instead because eSet objects enforce strict matrix dimensions and metadata alignment that complicate simple tidy-data workflows.

Data Requirements

• High-throughput data matrices (e.g., expression values) where rows represent features and columns represent samples.
• Sample covariates formatted as an AnnotatedDataFrame.
• Feature covariates formatted as an AnnotatedDataFrame.
• Normalization state: Typically normalized quantitative data (e.g., log-expression), though raw matrices can be stored.

Key Parameters

• assayData (assayDataNew()): High-throughput data stored as a list, environment, or lockedEnvironment containing identically sized matrices.
• phenoData (AnnotatedDataFrame()): Sample covariates matching the column names of assayData.
• featureData (AnnotatedDataFrame()): Feature covariates matching the row names of assayData.
• experimentData (MIAME()): Experimental description and metadata.
• annotation (character()): Label identifying the associated annotation package.

Best Practices

• Use accessor functions like exprs() and pData() to retrieve or assign data rather than accessing slots directly.
• Ensure that all matrices within assayData have identical row and column dimensions, and that featureNames and sampleNames match across all slots.
• Call validObject() after making structural modifications to an eSet to guarantee that the object remains internally consistent.
• Use storageMode() to check or set how assayData is stored, preferring lockedEnvironment to prevent accidental side-effects while maintaining memory efficiency.

Common Pitfalls

• Modifying a lockedEnvironment directly causes an error; fix this by extracting the element, modifying it, and reassigning it, or using replacement methods like exprs(obj) <- value.
• Variable shadowing occurs when applying functions with esApply; fix this by ensuring the applied function correctly references the covariate names populated from the pData dataframe.
• Objects become transiently invalid during manual slot modification; fix this by calling validObject() after updates to ensure dimensional consistency across assayData and phenoData.

Alternatives

• SummarizedExperiment: The modern Bioconductor standard for containerizing genomic data, supporting coordinate-based genomic ranges.
• oligo: Provides alternative, specialized class implementations for handling SNP and exon array data.
• MultiAssayExperiment: For integrating multiple different omics assays on the same set of biological specimens.

Citations

• R. Gentleman, V. Carey, M. Morgan, S. Falcon and H. Khan. "esApply Introduction".
• Martin T. Morgan and H. Khan. "Biobase development and the new eSet".

References

• Homepage: https://bioconductor.org/packages/Biobase
• Vignette: https://bioconductor.org/packages/release/bioc/vignettes/Biobase/inst/doc/ExpressionSetIntroduction.pdf

<!-- biomate-cta -->

---

Run this on BioMate

This skill is the knowledge layer — when, why, and how to use biobase. To run this analysis on your own data with managed compute, automated QC, and reproducible outputs, use BioMate — free to start.

▶ Open biobase on BioMate →