Scdrake: a reproducible and scalable pipeline for scRNA-seq data analysis

Area of Science:

• Genomics
• Bioinformatics
• Computational Biology

Background:

• Primary analysis of single-cell RNA-seq (scRNA-seq) data is established.
• Secondary analysis often relies on custom scripts, lacking standardization and reproducibility.
• There is a need for automated, reproducible pipelines in R for scRNA-seq secondary analysis.

Purpose of the Study:

• To develop a fully automated R pipeline for the secondary analysis of scRNA-seq data.
• To provide a reproducible, scalable, and efficient workflow adhering to best programming practices.
• To facilitate comprehensive secondary analysis, including quality control, normalization, and cell type annotation.

Main Methods:

• Developed scdrake, an R-based automated workflow using the drake framework.
• Integrated modules for quality control, filtering, normalization, dimensionality reduction, clustering, and differential expression analysis.
• Utilized a Docker image for straightforward setup and enhanced reproducibility.

Main Results:

• scdrake provides a fully automated secondary analysis workflow for scRNA-seq data.
• The pipeline is reproducible, scalable, and offers efficient execution with access to intermediate results.
• Outputs include comprehensive HTML reports and facilitate multi-sample integration and marker gene detection.

Conclusions:

• scdrake addresses the need for an automated and reproducible R pipeline for scRNA-seq secondary analysis.
• The tool enhances efficiency and accessibility of complex scRNA-seq data analysis.
• Availability via Docker and open-source code promotes widespread adoption and reproducibility.