Idéfix: identifying accidental sample mix-ups in biobanks using polygenic scores.

Area of Science:

• Bioinformatics
• Genetics
• Biobanking

Background:

• Identifying sample mix-ups in biobanks is crucial for accurate clinical pharmacogenetics.
• Existing methods are limited, especially in datasets lacking omics data.
• Relying solely on sex can only detect about half of all sample mix-ups.

Purpose of the Study:

• To introduce Idéfix, a new method for detecting accidental sample mix-ups in biobanks.
• To improve the reliability of genetic data for clinical applications by minimizing mix-up impact.
• To enhance the utility of biobank data for pharmacogenetic research.

Main Methods:

• Developed and applied the Idéfix method using polygenic scores (PGSs) for 25 traits in 32,786 participants from the Lifelines biobank.
• Compared actual phenotypes with calculated PGSs to identify discordance indicative of sample mix-ups.
• Validated Idéfix performance through simulations with induced mix-ups, assessing its accuracy (AUC).

Main Results:

• Idéfix achieved an AUC of 0.90 in simulations using 25 PGSs and sex, a significant improvement over sex alone (AUC 0.75).
• The method can identify a high-quality subset of participants (34.4% in Lifelines) with a very low probability of mix-ups.
• Application in Lifelines reduced the sample mix-up rate from 0.15% to 0.01%.

Conclusions:

• Idéfix offers a robust and scalable solution for identifying sample mix-ups in biobanks.
• The method's performance is expected to improve with more powerful Genome-Wide Association Studies (GWASs).
• Idéfix enables the reliable use of genetic data for clinical pharmacogenetics by ensuring sample integrity.