Abstract #174
Section: Breeding and Genetics (orals)
Session: Breeding and Genetics II: Methodologies, Inbreeding and Breeding Strategies
Format: Oral
Day/Time: Monday 4:45 PM–5:00 PM
Location: Room 301 B
Session: Breeding and Genetics II: Methodologies, Inbreeding and Breeding Strategies
Format: Oral
Day/Time: Monday 4:45 PM–5:00 PM
Location: Room 301 B
# 174
Integrating genomic information and large-scale FTIR-based phenotyping for the genetic improvement of cheese-making traits in Brown Swiss cattle.
Francesco Tiezzi*1, Christian Maltecca1, Hugo Toledo Alvarado3, Attilio Rossoni2, Enrico Santus2, Giovanni Bittante3, Alessio Cecchinato3, 1Department of Animal Science, North Carolina State University, Raleigh, NC, 2Italian Brown Swiss Breeders' Association, Bussolengo, Italy, 3Department of Agronomy, Food, Natural resources, Animals and Environment, Legnaro, Padova, Italy.
Key Words: genomic selection, FTIR phenotyping, cheese yield
Integrating genomic information and large-scale FTIR-based phenotyping for the genetic improvement of cheese-making traits in Brown Swiss cattle.
Francesco Tiezzi*1, Christian Maltecca1, Hugo Toledo Alvarado3, Attilio Rossoni2, Enrico Santus2, Giovanni Bittante3, Alessio Cecchinato3, 1Department of Animal Science, North Carolina State University, Raleigh, NC, 2Italian Brown Swiss Breeders' Association, Bussolengo, Italy, 3Department of Agronomy, Food, Natural resources, Animals and Environment, Legnaro, Padova, Italy.
The objective of this study was to evaluate genotyping and phenotyping strategies for the improvement of cheesemaking traits in Italian Brown cattle. Phenotyping was considered as model-cheese manufacturing based (high-cost, low-throughput) or Fourier-transform infrared spectroscopy enabled (low cost, high throughput). Data were from 1,011 cows phenotyped using the model-cheese manufacturing method (considered as LAB in the present study) and included 3 cheese yield traits (%CY: fresh curd, curd total solids, and curd water as a percent of the weight of the processed milk), 4 curd nutrient recovery traits (REC: fat, protein, total solids, and the energy of the curd as a percent of the same nutrient in the processed milk). The FIELD data set consisted of ~660,000 FTIR predictions for the same traits on ~35,000 cows. Pool of genotyped individuals consisted of: 1,011 LAB cows, 1,493 of the FIELD cows, 181 sires with LAB and FIELD daughters (siresA), 540 sires with FIELD daughters (siresB). Individuals were genotyped with different SNP panels and imputed to 50k. A 4-fold cross-validation was used to assess predictive ability of models, meant as the ability to predict masked LAB records from daughters of progeny testing bulls. The correlation between observed and predicted LAB measures in validation was averaged over the 4 training-validation sets. Sets of phenotypic information were so defined: M1, LAB cows from the training set; M2, LAB+FIELD cows from the training set; M3, LAB cows in training and all FIELD cows. M2 and M3 considered LAB and FIELD cows as distinct traits. As for the genomic information, sets were defined as: no individuals genotyped; SiresA; SiresA+LAB; SiresA+LAB+SiresB; SiresA+LAB+SiresB+FIELD. For each trait, a total of 3 × 5 = 15 models were implemented. Predictions were obtained using the Single-Step GBLUP method. Results show that the use of genomic information does not provide any advantage in predictive ability. Prediction models that included FIELD records showed an advantage for the traits RECenergy, RECprotein, RECsolids and CYsolids over the models that included LAB records only. For RECfat, CYwater and CYcurd, this advantage was negligible.
Key Words: genomic selection, FTIR phenotyping, cheese yield
