simuRare

simuRare simulates genomic datasets containing rare and common single nucleotide polymorphisms (SNPs) by integrating logistic regression–based imputation with a resampling approach using 1000 Genomes Project reference data to preserve linkage disequilibrium and minor allele frequency for evaluation of statistical methods and disease penetrance models.

Key Features:

• Regression-Based Imputation: Uses logistic regression models with 1000 Genomes Project reference data to predict rare variant presence from nearby common variants in SNP array datasets.
• Resampling Approach: Generates simulated samples via resampling that maintain linkage disequilibrium (LD) and allele frequency distributions observed in reference data.
• Preservation of Genomic Properties: Retains sample characteristics such as LD and minor allele frequency to produce realistic genetic structures for method evaluation.

Scientific Applications:

• Evaluate Statistical Methods: Enables benchmarking and evaluation of statistical and bioinformatics methods under realistic genetic architectures containing rare and common SNPs.
• Study Disease Mechanisms: Allows imposing known disease penetrance models on simulated data to investigate the impact of rare variants on Mendelian and common (complex) diseases.
• Enhance Computational Efficiency: Combines regression-based imputation with resampling to provide computationally efficient simulation of large-scale genetic datasets while preserving key genomic properties.

Methodology:

Logistic regression models for imputation using 1000 Genomes Project reference data; resampling to generate simulated samples that preserve linkage disequilibrium and minor allele frequency; capability to impose known disease penetrance models.