CORESH: a gene signature-based search engine for public gene expression datasets

Area of Science:

• Bioinformatics
• Computational Biology
• Genomics

Background:

• Public data repositories like Gene Expression Omnibus (GEO) offer vast experimental data.
• Current search methods for GEO datasets rely on keywords and high-quality metadata, limiting hypothesis-driven discovery.
• Researchers often need to find transcriptional similarity between their data and public datasets to infer similar biological processes.

Purpose of the Study:

• To introduce CORESH, a web server for data-driven discovery of GEO datasets.
• To enable researchers to find datasets matching a user-provided gene signature in a systematic manner.
• To facilitate the identification of experimental conditions associated with specific gene expression patterns.

Main Methods:

• CORESH utilizes a compendium of over 40,000 human and 40,000 mouse datasets from GEO.
• The web server accepts a gene signature (e.g., upregulated genes) as input.
• It outputs a ranked list of datasets based on similar expression patterns of the input genes.

Main Results:

• CORESH systematically identifies GEO datasets with matching gene expression patterns.
• The tool enables a data-driven approach to finding relevant public datasets.
• Discovered datasets can reveal experimental conditions linked to the query signature.

Conclusions:

• CORESH provides a valuable resource for researchers seeking to leverage public gene expression data.
• The server aids in uncovering underlying biological mechanisms and guiding experimental validation.
• CORESH is freely accessible and regularly updated with the latest GEO data.