pipeComp

pipeComp evaluates computational pipelines for single-cell RNA sequencing (scRNAseq) by benchmarking methods across filtering, doublet detection, normalization, feature selection, denoising, dimensionality reduction, and clustering to assess their impact on downstream analysis.

Key Features:

• Multi-Level Evaluation Metrics: Employs a multi-level evaluation approach to assess both intermediate and final stages of scRNAseq analysis, quantifying how each tool affects cell population recovery and clustering performance.
• Comprehensive Pipeline Benchmarking: Benchmarks scRNAseq pipelines using simulated and real datasets with known cell identities and covers filtering, doublet detection, normalization, feature selection, denoising, dimensionality reduction, and clustering.
• Extensible Framework: Allows integration of additional steps, tools, or evaluation metrics to extend benchmarking to new methods and other bioinformatics workflows.
• Practical Recommendations: Produces systematic recommendations and proposed pipeline configurations based on comparative evaluations to optimize scRNAseq data processing.
• Impact Analysis: Includes analyses of the impact of removing unwanted variation on differential expression analysis.

Scientific Applications:

• scRNAseq pipeline selection and optimization: Supports selection and optimization of computational pipelines for single-cell RNA sequencing data processing and analysis.
• Cellular heterogeneity and biological process analysis: Enables robust comparisons that aid studies of cellular heterogeneity and complex biological processes at the single-cell level.

Methodology:

Applies multi-purpose, multi-level evaluation metrics to assess tools within scRNAseq pipelines on simulated and real datasets with known cell identities.