SCUBA implements a storage format-agnostic API for single-cell data access in R.

Area of Science:

• Computational Biology
• Bioinformatics
• Genomics

Background:

• Single-cell data analysis tools in R and Python lack interoperability.
• Integrating tools across ecosystems requires significant programming expertise.
• Object format conversion is error-prone and memory-intensive.

Purpose of the Study:

• To develop a unified data access API for common R and Python single-cell object formats.
• To simplify and improve the reproducibility of single-cell data analysis.
• To reduce the complexity of integrating diverse analysis tools.

Main Methods:

• Developed SCUBA (Single-Cell Unified Backend API) as an R package.
• Extended Seurat's data access approach to SingleCellExperiment and anndata objects.
• Implemented new data-specific access functions for supported object types.

Main Results:

• SCUBA provides a unified API for R and Python single-cell object formats.
• Performance scales effectively across all supported formats.
• SCUBA extracts only necessary data, leaving original objects unmodified, reducing errors and memory usage.
• Code written with SCUBA supports multiple object types with consistent syntax.

Conclusions:

• SCUBA enhances interoperability between R and Python single-cell analysis packages.
• Adoption of SCUBA can improve collaboration and reproducibility in the field.
• SCUBA offers a simpler, less error-prone alternative to object conversion for data analysis and visualization.