Singe cell RNA sequencing data processing using cloud-based serverless computing

Area of Science:

• Computational Biology
• Bioinformatics
• Cloud Computing

Background:

• Single-cell RNA sequencing (scRNA-seq) is crucial for understanding cellular activity.
• Processing large scRNA-seq datasets demands substantial computational resources.
• Cloud computing offers scalable solutions without significant infrastructure investment.

Purpose of the Study:

• To develop a novel, generalizable methodology for accelerating computationally intensive workflows using serverless cloud computing.
• To create an on-demand computational environment using serverless functions for biological data analysis.

Main Methods:

• Leveraging serverless cloud functions to create automatically provisioned computation units.
• Optimizing a scRNA-seq workflow by integrating serverless functions.
• Testing the methodology on public peripheral blood mononuclear cell (PBMC) datasets and a large NIH MorPhiC dataset (450 GB).

Main Results:

• Demonstrated major speedups in processing large scRNA-seq datasets compared to traditional workflows.
• Successfully processed a 450 GB human scRNA-seq dataset across 86 cell lines.
• Validated the cost-efficiency and performance of serverless computing for scRNA-seq analysis.

Conclusions:

• Serverless cloud computing provides a cost-effective and efficient solution for accelerating scRNA-seq data processing.
• The proposed methodology is generalizable for computationally intensive bioinformatics workflows.
• This approach enables rapid deployment of computational power for large-scale biological data analysis.