How far are we from the era of big data in transcriptomics? Lessons from the bacterial data in GEO

Area of Science:

• Functional genomics
• Bioinformatics
• Data science

Background:

• The Gene Expression Omnibus (GEO) is a major repository for functional genomics data, housing millions of entries from microarrays and RNA-sequencing (RNA-seq).
• Bacterial transcriptomic data within GEO holds significant potential for large-scale meta-analysis, particularly in systems biology, due to the vast diversity of biological conditions represented.
• Despite the rise of RNA-seq, bacterial microarrays constitute a substantial portion (~48%) of GEO entries, necessitating their re-evaluation and improved accessibility.

Purpose of the Study:

• To assess the current status of bacterial microarray and RNA-seq data and associated metadata within the GEO repository.
• To identify inconsistencies in GEO metadata and community data usage that impede automated access and interpretation of high-throughput analyses.
• To investigate the availability and processability of bacterial microarray data for large-scale reanalysis.

Main Methods:

• Systematic evaluation of bacterial transcriptomic data and metadata quality in GEO.
• Analysis of data format standardization and community usage patterns.
• Assessment of challenges in microarray data processing and normalization for integration into large-scale reanalysis.

Main Results:

• Identified significant inconsistencies in GEO metadata documentation and user-generated data, compromising automated data retrieval and biological context.
• Microarray data processing and normalization pose challenges, limiting its integration into large-scale reanalysis efforts.
• Lack of standardized formats restricts the reusability of at least 44% of approximately 45,000 bacterial microarray entries in GEO.

Conclusions:

• GEO transcriptomic data and metadata are valuable but require continuous maintenance and revision.
• Addressing inconsistencies and standardizing formats are crucial for unlocking the full potential of GEO data for big data initiatives.
• Proposed guidelines aim to improve the Findability, Accessibility, Interoperability, and Reusability (FAIR principles) of GEO data for enhanced scientific discovery.