Automated tetraploid genotype calling by hierarchical clustering

Area of Science:

• Genetics
• Bioinformatics
• Plant Breeding

Background:

• SNP arrays are crucial for autotetraploid research but require accurate genotype calling.
• Calibrating signal intensity to allele dosage is essential for marker utility.

Purpose of the Study:

• To develop an automated computational method for accurate tetraploid genotype calling.
• To provide an R package, ClusterCall, for efficient SNP data analysis in autotetraploids.

Main Methods:

• Utilizing hierarchical clustering and multiple F1 populations for signal intensity-to-dosage calibration.
• Employing a training phase with F1 populations and a prediction phase with diverse sets.
• Implementing a concordance metric to assess marker reliability and compare algorithms.

Main Results:

• ClusterCall achieved higher concordance (94.6%) than fitTetra (82.5%) across three potato families.
• Enabled genotype calling for 5218 SNPs in the SolCAP diversity panel, significantly more than previous manual methods.
• Identified a significant association for vine maturity using newly called markers.

Conclusions:

• ClusterCall provides an efficient and accurate method for autotetraploid genotype calling.
• The R package facilitates the use of SNP data in autotetraploid research and breeding.
• Improved genotype calling enhances genetic discoveries and breeding applications.