splatter

splatter simulates single-cell RNA sequencing (scRNA-seq) count data to generate realistic datasets for method development and validation.

Key Features:

• Reproducibility: Supports specification of simulation parameters to produce reproducible scRNA-seq count datasets.
• Multiple simulation methods: Includes the Splat method, which uses a gamma-Poisson distribution, and supports scenarios such as single cell populations, multiple cell types, and differentiation paths.
• Parameter estimation: Enables estimation of simulation parameters directly from real scRNA-seq datasets to match empirical distributions.
• Comparison functions: Provides functions to compare real and simulated datasets for assessment of simulation fidelity.
• Bioconductor integration: Implemented within the Bioconductor ecosystem.

Scientific Applications:

• Method development and benchmarking: Generating realistic count data for development and benchmarking of scRNA-seq analysis algorithms.
• Cell differentiation modeling: Simulating differentiation paths to evaluate trajectory and lineage inference methods.
• Pipeline validation and artifact detection: Assessing potential artifacts and biases in data analysis pipelines by comparing simulated and empirical data.
• Experimental design optimization: Simulating expected outcomes under varying scenarios to inform experimental design choices.

Methodology:

The Splat method implements a gamma-Poisson statistical model to generate scRNA-seq count data, with parameters that can be estimated from real datasets to capture gene expression variability.