Genotype calling in tetraploid species from bi-allelic marker data using mixture models

Area of Science:

• Genetics
• Bioinformatics
• Plant Breeding

Background:

• Automated genotype calling for tetraploid species was previously not feasible, hindering genetic analysis.
• Existing software for diploid species cannot accommodate the five possible genotypes in tetraploids (aaaa to bbbb).

Purpose of the Study:

• To develop a novel algorithm and R package, fitTetra, for automated genotype calling in tetraploid species.
• To enable accurate genetic analysis in tetraploids using intensity data from genotyping assays.

Main Methods:

• Developed an algorithm implemented in the R package fitTetra.
• Utilized mixture models with five components, each representing a possible tetraploid genotype.
• Incorporated constraints for Hardy-Weinberg equilibrium (HWE) ratios in model fitting.

Main Results:

• The fitTetra package successfully assigns genotypes for bi-allelic markers in tetraploid samples.
• The software identifies and rejects unreliable markers and assigns missing scores to ambiguous genotypes.
• Validated with potato varieties using an Illumina SNP array, showing reliable genotype assignments.

Conclusions:

• fitTetra provides a robust solution for automated genotype calling in tetraploid species.
• The software facilitates genetic analysis by accurately determining genotypes from intensity signals.
• The R package fitTetra is freely available for use in genetic research.