Consistently processed RNA sequencing data from 50 sources enriched for pediatric data

Area of Science:

• Genomics
• Bioinformatics
• Cancer Research

Background:

• Tumor gene expression analysis requires large cohorts for statistical power.
• Inconsistent data processing and metadata accuracy hinder the analysis of diverse datasets.
• Existing datasets often suffer from variability, limiting their utility for robust research.

Purpose of the Study:

• To develop consistently processed gene expression compendia from a large number of diverse RNA sequencing datasets.
• To harmonize clinical metadata alongside gene expression values for comprehensive analysis.
• To provide unrestricted access to high-quality, standardized datasets for cancer research.

Main Methods:

• Acquired RNA sequencing data from public repositories and clinical partners.
• Assessed data quality, quantified gene expression, and harmonized clinical metadata.
• Developed a dockerized pipeline for consistent data processing and analysis.

Main Results:

• Generated five compendia comprising 16,446 RNA sequencing datasets.
• Ensured consistent processing and harmonized metadata across all datasets.
• Released expression values and metadata without access restrictions.

Conclusions:

• The developed compendia provide a powerful, standardized resource for tumor gene expression analysis.
• These datasets facilitate diverse research applications, including tumor type comparison, cell line validation, and n-of-1 studies.
• The freely available pipeline enables the comparison of new data to the compendia, advancing cancer genomics research.