Estimating autozygosity from high-throughput information: effects of SNP density and genotyping errors

Area of Science:

• Genetics
• Animal Breeding

Background:

• Runs of homozygosity (ROH) estimate inbreeding (autozygosity) using SNP genotypes.
• Empirical ROH identification is affected by SNP chip density and genotyping error tolerance.

Purpose of the Study:

• To analyze the impact of SNP chip density and genotyping errors on autozygosity estimation using ROH.
• To compare ROH analysis between a high-density SNP chip (777,972 SNPs) and a lower-density chip (50k) in cattle.

Main Methods:

• Analysis of ROH in three cattle populations using genotype data from two SNP chips of varying densities.
• Evaluation of how allowing heterozygous calls affects ROH identification and autozygosity estimation.

Main Results:

• Lower-density (50k) SNP chips overestimate short ROH (<4 Mb) and underestimate long ROH (>8 Mb).
• Higher-density SNP chips underestimate long ROH unless minor heterozygous calls are permitted.
• Current software struggles to differentiate true ROH breaks from scattered heterozygous calls.

Conclusions:

• SNP chip density and genotyping errors introduce bias in autozygosity estimation via ROH.
• Standard ROH methods for humans are inadequate for livestock due to higher autozygosity.
• Conservative autozygosity predictions for ROH >4 Mb are obtained with 50-60k SNP chips.