RAREsim2: Flexible simulation of rare variant genetic data using real haplotypes.

Area of Science:

• Genetics
• Bioinformatics
• Computational Biology

Background:

• Realistic simulated genetic data is essential for advancing genetic research methods and study designs.
• Existing simulation tools often struggle to accurately model rare variants and incorporate key genetic information like functional annotations and linkage disequilibrium.

Purpose of the Study:

• Introduce RAREsim2, an updated rare-variant simulation algorithm.
• Provide streamlined software with new functionalities for simulating individual and variant-level differences to represent diverse causal models.

Main Methods:

• RAREsim2 simulates genetic data using real genetic haplotypes.
• The tool incorporates individual-level differences (e.g., cases vs. controls, batch effects) and variant-level differences for various causal models.
• Demonstrated utility with Burden, SKAT, and SKAT-O rare variant association methods across multiple simulation scenarios.

Main Results:

• RAREsim2 successfully maintained Type I Error rates across simulations.
• The optimal rare variant association test performance matched known patterns (Burden, SKAT, SKAT-O).
• Highlighted capabilities in simulating diverse genetic ancestries, gene sizes, association strengths, and risk variant proportions.

Conclusions:

• RAREsim2 offers enhanced flexibility and ease of use for simulating realistic genetic scenarios.
• The updated algorithm improves the ability to model complex genetic data, aiding in methodological development and study design.
• Facilitates the simulation of genetic regions with known variant functions and disease associations.