DupChecker: a bioconductor package for checking high-throughput genomic data redundancy in meta-analysis

Area of Science:

• Genomics
• Bioinformatics

Background:

• Meta-analysis is crucial for high-throughput genomic data analysis, enhancing the power to detect biological signals.
• Sample duplication in public databases, particularly for gene expression data, is a common issue.
• Unidentified duplicates can cause false positives, skewed clustering, and model overfitting.

Purpose of the Study:

• To introduce DupChecker, a Bioconductor package designed for efficient identification of duplicated samples in genomic datasets.
• To provide a practical tool for researchers to ensure data integrity before meta-analysis.

Main Methods:

• Development of the DupChecker package utilizing MD5 fingerprinting of raw data.
• Implementation of a robust method for detecting sample duplication.

Main Results:

• DupChecker efficiently identifies duplicated samples through MD5 fingerprint generation.
• A demonstration using real data showcases the package's functionality and output.

Conclusions:

• Failure to address sample duplication can compromise the validity of meta-analysis results.
• It is recommended to apply DupChecker to all gene expression datasets prior to analysis to prevent data contamination.