fitTetra

fitTetra fits mixture models to allele signal ratios to call genotypes and estimate allele dosages in tetraploid organisms.

Key Features:

• Mixture model fitting: Fits a mixture model to dosage signal distributions and estimates parameters separately for each sub-collection.
• Parental genotype integration: Incorporates parental genotypes from full-sib families to predict expected segregation patterns in allele dosages.
• Population-specific mixing proportion estimation: Estimates mixing proportions using parental dosages and inheritance models for full-sib families and assumes Hardy-Weinberg equilibrium for unstructured populations.
• Signal ratio distribution assumption: Assumes consistent signal ratio distributions of dosage classes across populations to allow borrowing parental data between subpopulations.
• Allele dosage prediction: Produces allele dosage calls for tetraploid SNPs informed by mixture-model parameters and parental segregation expectations.

Scientific Applications:

• High-throughput genotyping of tetraploid species: Applied to array data, including testing on an Affymetrix Axiom 60 k array dataset.
• Genotype calling in full-sib families: Uses parental genotypes and inheritance models to improve dosage calls in family-based populations.
• Genotype calling in unstructured populations: Estimates genotype proportions under Hardy-Weinberg equilibrium for population-based studies.
• Improved SNP call recovery: Demonstrated increased SNP call yield compared with previous versions and tools such as ClusterCall, reporting at least ~20% more SNPs called in tested datasets.

Methodology:

Fit a mixture model to dosage signal ratios with parameter estimation per sub-collection; use parental genotypes to predict segregation-based allele dosages and derive mixing proportions for full-sib families; assume Hardy-Weinberg equilibrium to estimate mixing proportions in unstructured populations; assume consistent signal ratio distributions across populations to borrow parental data for fitting.