TidyGEO: preparing analysis-ready datasets from Gene Expression Omnibus

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• The Gene Expression Omnibus (GEO) is a large public repository of high-throughput gene expression data.
• GEO data requires significant cleaning and reformatting of sample-level annotations for effective secondary analysis.
• Inconsistent annotation structures across GEO series hinder data usability.

Purpose of the Study:

• To develop TidyGEO, a web-based tool to automate the tidying and reformatting of GEO sample annotations.
• To facilitate secondary research by making GEO data more accessible and analyzable.
• To address the challenges of manual data cleaning and the need for computational expertise.

Main Methods:

• TidyGEO offers functionalities for selecting, renaming, splitting, merging, standardizing, and filtering annotation columns.
• The tool supports integrating annotations with assay data and restructuring assay data.
• TidyGEO generates reproducible code for data processing steps.

Main Results:

• TidyGEO provides essential data-cleaning tasks for sample-level annotations from GEO.
• Users can efficiently prepare GEO data for downstream analysis.
• The tool enhances the reproducibility of data analysis.

Conclusions:

• TidyGEO effectively addresses the challenges of processing GEO data for secondary research.
• The tool empowers researchers with varying computational expertise to analyze complex biological datasets.
• TidyGEO promotes efficient and reproducible use of public gene expression data.