Improving power for rare-variant tests by integrating external controls.

Area of Science:

• Genomics and Bioinformatics
• Statistical Genetics
• Computational Biology

Background:

• Decreasing sequencing costs are rapidly increasing the number of available sequenced genomes.
• External control samples offer potential to enhance the power of rare-variant association tests.
• Batch effects from diverse sequencing platforms and analysis pipelines can inflate type I error rates when using external controls.

Purpose of the Study:

• To develop novel statistical methods for integrating external control samples into rare-variant tests.
• To control for type I error rates introduced by batch effects when using external controls.
• To improve the power of single, gene-based, and region-based rare-variant tests.

Main Methods:

• Proposed novel summary statistics-based rare-variant tests.
• Developed an approach to assess batch effects by comparing odds ratio estimates from internal versus combined internal and external controls.
• Implemented methods for single, gene-based, and region-based analyses.

Main Results:

• The proposed methods effectively control for type I error rates when integrating external controls.
• The novel approach demonstrated a substantial improvement in statistical power.
• Validated through simulation experiments and analyses of age-related macular degeneration and type 2 diabetes datasets.

Conclusions:

• Novel summary statistics-based tests allow effective integration of external controls in rare-variant analyses.
• The methods successfully control for type I error, mitigating batch effect inflation.
• This approach offers a powerful strategy for increasing the discovery potential of rare genetic variants in large-scale genomic studies.