Inferring unknown genotypes of sires at codominant deoxyribonucleic acid markers in half-sib families

Deoxyribonucleic acid from sires is usually not available from experiments aimed at QTL mapping for traits of the dam in cow-calf operations and free range sheep populations. In this study, methods to reconstruct sire genotypes using genotype information from large half-sib progeny were developed. The methods are based on 1) all offspring genotypes are compatible with more than 1 genotype for the sire, but 1 of the genotypes is more likely than the others when comparing the proportion of the different genotypes among offspring with its expected values assuming Mendelian inheritance, or 2) all offspring genotypes are compatible with just 1 possible genotype for the sire in the pedigree. A Monte Carlo simulation experiment was carried out to test the methods with 1 million replicates. A 99.7% correct sire genotype reconstruction was obtained with 30 offspring and a DNA marker with 3 or more alleles segregating at similar frequencies. Methods to test for incorrect paternity in half-sib offspring without DNA from the sire were also developed. A maximum likelihood method was developed to test for departure of Mendelian segregation due to a contaminating sire whose offspring are fully compatible with the genotype of the pedigree sire. A large number of offspring was needed to detect offspring from a contaminating sire (1,000 progeny for a power of 0.99 and proportion of true paternity of the pedigree of 0.80). Multi-marker methods were also developed for detection of paternity misidentification. Probabilities of detection of wrong paternity for a contaminating sire not sharing any alleles with the sire in the pedigree were 0.95 and 0.99 when using 5 and 10 markers in 30 half-sib offspring, respectively. The methods to infer the sire genotypes were tested with 49 progeny of a Merino ram whose genotype was inferred for 7 microsatellites. Methods to infer genotype of the sire are feasible, but QTL mapping experiments without DNA from the sires are more costly due to the need of genotyping markers in progeny for which the sire in the pedigree is homozygous.