Modular, efficient and constant-memory single-cell RNA-seq preprocessing.

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Single-cell RNA sequencing (scRNA-seq) generates high-dimensional data.
• Efficient and accurate preprocessing is crucial for downstream analysis.
• Existing workflows may face scalability or performance limitations.

Purpose of the Study:

• To introduce a novel preprocessing workflow for scRNA-seq data.
• To balance computational efficiency with analytical accuracy.
• To demonstrate the workflow's flexibility for advanced applications like RNA velocity.

Main Methods:

• Development of a modular preprocessing pipeline.
• Integration of the kallisto and bustools software.
• Optimization for speed and constant memory usage.
• Demonstration of scalability for large datasets.

Main Results:

• The workflow achieves near-optimal speed for scRNA-seq data preprocessing.
• It exhibits constant memory requirements, ensuring scalability.
• The modular design allows for adaptation to various analytical needs.
• Successful application in RNA velocity analyses was demonstrated.

Conclusions:

• The kallisto-bustools based workflow offers an efficient and accurate solution for scRNA-seq preprocessing.
• Its scalability and flexibility make it suitable for large-scale genomic studies.
• This method facilitates advanced analyses such as RNA velocity.