pyVIPER: a fast and scalable Python package for protein activity estimation and master regulator analysis of...

Area of Science:

• Computational Biology
• Genomics
• Bioinformatics

Background:

• Single-cell RNA sequencing (scRNA-seq) provides high-resolution insights into cellular heterogeneity.
• Low signal-to-noise ratios in scRNA-seq data hinder quantitative analyses.
• Gene regulatory network (GRN) analysis, like the VIPER algorithm, helps elucidate cellular state determinants by identifying Master Regulator proteins.

Purpose of the Study:

• To address the need for scalable tools for analyzing large scRNA-seq datasets.
• To introduce pyVIPER, a Python-based tool for protein activity inference from transcriptional data.
• To improve the efficiency and accessibility of GRN analysis in single-cell research.

Main Methods:

• Developed pyVIPER, a Python toolkit for protein activity inference.
• Implemented flexible data transformation/postprocessing modules and enrichment analysis algorithms.
• Integrated a novel data structure for GRN manipulation and compatibility with scverse and scanpy.
• Leveraged PyTorch-based GPU acceleration and optimized core operations for enhanced performance.

Main Results:

• pyVIPER demonstrates orders-of-magnitude improvements in runtime efficiency compared to R-based VIPER.
• Analysis time for large scRNA-seq datasets is reduced from hours to minutes.
• The tool is memory-efficient and highly scalable for datasets up to hundreds of thousands of cells.

Conclusions:

• pyVIPER is a fast, memory-efficient, and scalable Python toolkit for protein activity inference in large-scale scRNA-seq data.
• Its hardware acceleration and scalability enable high-throughput VIPER-based analysis for diverse single-cell datasets.
• pyVIPER facilitates integration with Python-based machine learning workflows, expanding mechanistic GRN analysis in single-cell research.