Rare variant testing of imputed data: an analysis pipeline typified.

Area of Science:

• Genetics
• Bioinformatics
• Statistical Genetics

Background:

• Rare variant association testing is crucial for understanding complex diseases.
• Existing methods often rely on whole-exome sequencing data.
• Imputed data offers a cost-efficient alternative for large-scale genetic studies.

Purpose of the Study:

• To develop and validate a pipeline for rare variant analysis using imputed genetic data.
• To provide quality control criteria and strategies for constructing analysis bins.
• To implement and integrate various rare variant testing methods within the pipeline.

Main Methods:

• Proposed a novel pipeline for rare variant analysis of imputed data.
• Developed quality control criteria and binning strategies for whole-genome analysis.
• Implemented burden tests (COLL, CMAT) and regression tests (REG, FRACREG, COLLREG) with the variable threshold (VT) approach.
• Integrated kernel tests (SKAT/SKAT-O) into the analysis framework.
• Applied the pipeline to a genome-wide association study of Alzheimer's disease.

Main Results:

• The pipeline demonstrated valid significance testing with controlled type I error rates.
• Strong association signals were detected near the APOE locus, indicating statistical power.
• Several suggestive rare variant associations were identified for regions including MCPH1, MED18, and NOTCH3.

Conclusions:

• The developed pipeline offers a straightforward, cost-efficient method for rare variant analysis in imputed data.
• This strategy complements next-generation sequencing approaches in rare variant studies.
• The findings support the feasibility and validity of using imputed data for rare variant association testing.