bigSCale: an analytical framework for big-scale single-cell data.

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) generates massive datasets, posing computational challenges for analysis.
• Existing scRNA-seq tools struggle with large data volumes, impacting processing time and marker gene identification sensitivity.

Purpose of the Study:

• To introduce bigSCale, a scalable analytical framework designed for analyzing millions of cells.
• To address limitations in current tools for handling large scRNA-seq datasets and improve sensitivity for marker gene discovery.

Main Methods:

• bigSCale employs a numerical model of noise using large sample sizes to handle scRNA-seq data sparsity.
• Utilizes a directed convolution strategy for processing large datasets while preserving single-cell transcript information.
• Incorporates modules for differential expression analysis, cell clustering, and marker identification.

Main Results:

• bigSCale demonstrated speed and accuracy in differential expression analysis on simulated and biological datasets.
• Analysis of 1.3 million mouse forebrain cells revealed improved resolution in cell clustering.
• Identified rare cell populations, including reelin-positive Cajal-Retzius neurons, with previously unrecognized heterogeneity.

Conclusions:

• bigSCale offers a scalable solution for analyzing large-scale scRNA-seq data.
• The framework enhances the identification of cellular heterogeneity and rare cell populations.
• Provides a robust tool to meet the growing computational demands of single-cell genomics.