GIGI-Check

GIGI-Check detects Mendelian-consistent genotyping errors in dense marker datasets from pedigrees to improve the accuracy of genetic analyses.

Key Features:

• Implementation: Implemented in C++ for computational analysis of pedigree genotype data.
• Detection of Mendelian-consistent (MC) errors: Targets Mendelian-consistent genotyping errors in dense variant datasets, including single nucleotide polymorphisms (SNPs) and sequencing data, in contrast to methods that focus on Mendelian-inconsistent (MI) errors.
• Scalability with large pedigrees: Engineered to handle large pedigrees efficiently for studies with many family members and complex inheritance patterns.
• Inheritance vector (IV) sampling: Employs a Markov chain Monte Carlo-based sampler to generate IVs using a moderately sparse yet informative set of markers.
• A1 statistic: Measures the percentage of sampled IVs that are inconsistent with observed genotypes to flag potential errors.
• A2 statistic: Evaluates the posterior probability of error configurations to provide a probabilistic assessment of genotyping errors.
• Sensitivity and evaluation: Simulation results show sensitivity near theoretical best-case scenarios even using the A1 statistic, with effectiveness evaluated across genotype patterns, marker density, error rates, allele frequencies, and number of sampled IVs.

Scientific Applications:

• Genetic analysis: Reduces false results due to genotyping errors in high-throughput sequencing and dense marker studies to improve reliability of downstream genetic analyses.
• Pedigree studies: Supports integrity and accuracy of inheritance analyses in large pedigree-based investigations.

Methodology:

Implemented in C++ and using a Markov chain Monte Carlo sampler to generate inheritance vectors from a moderately sparse marker set, GIGI-Check computes the A1 statistic (percent of sampled IVs inconsistent with observed genotypes) and the A2 statistic (posterior probability of error configurations) and its performance was assessed by simulations varying genotype patterns, marker density, error rates, allele frequencies, and number of sampled IVs.