ALTHOUGH GENOMIC SELECTION HAS SUBSTANTIALLY INCREASED RATES OF GENETIC GAIN IN LIVESTOCK, SEVERAL CHALLENGES REMAIN, INCLUDING GENOTYPE BY ENVIRONMENTAL INTERACTIONS AND NON-ADDITIVE GENETIC EFFECTS. CURRENT GENOMIC PREDICTION METHODS ARE BASED ON EMPIRICAL LINEAR STATISTICAL MODELS OF RELATIONSHIPS BETWEEN GENOTYPES AND PHENOTYPES, WHICH TYPICALLY PROVIDE USEFUL PREDICTIONS FOR THE CONDITIONS (GENETIC AND ENVIRONMENTAL) THAT PREVAIL IN THE TRAINING DATA BUT SUFFER FROM POOR EXTRAPOLATION TO OTHER CONDITIONS. THE USE OF THESE BLACK BOX APPROACHES IS DESPITE THE AVAILABILITY OF SUBSTANTIAL KNOWLEDGE AND ASSOCIATED MECHANISTIC MODELS OF THE PHYSIOLOGICAL AND BIOLOGICAL PROCESSES THAT UNDERLY ANIMAL PERFORMANCE UNDER DIFFERENT CONDITIONS, INCLUDING HEAT AND COLD STRESS. OUR HYPOTHESIS IS THAT THESE SHORTCOMINGS CAN BE OVERCOME BY INTEGRATING THESE MECHANISTIC MODELS INTO METHODS FOR GENOMIC PREDICTION, AS WAS RECENTLY DEMONSTRATED USING A CROP GROWTH MODEL IN MAIZE. THE GOAL OF THE PROPOSED WORK IS TO DEVELOP, IMPLEMENT, AND EVALUATE THIS INTEGRATION IN THE CONTEXT OF GROW-FINISH PIGS. WE WILL CAPITALIZE ON ALREADY EXISTING MECHANISTIC PIG GROWTH MODELS AND THE ADVANCED BAYESIAN HIERARCHICAL GENERALIZED LINEAR GENOMIC PREDICTION METHODS THAT WERE DEVELOPED FOR MAIZE, WHICH WILL BOTH NEED SUBSTANTIAL UPDATING AND ADAPTATION FOR THIS APPLICATION. INTEGRATION OF BIOLOGICAL MODELS INTO GENOMIC EVALUATION IS EXPECTED TO HAVE SUBSTANTIAL IMPACTS ON PHENOTYPE RECORDING AND BREEDING PROGRAMS ACROSS ENVIRONMENTS AND POPULATIONS, RESULTING IN MORE ROBUST AND EFFECTIVE GENETIC IMPROVEMENT. WE HAVE ESTABLISHED A MULTI-DISCIPLINARY RESEARCH TEAM OF EXPERTS IN PIG NUTRITION AND GROWTH MODELING, MATHEMATICAL AND STATISTICAL QUANTITATIVE GENETICS, AND SWINE BREEDING. WE WILL HAVE ACCESS TO EXTENSIVE PHENOTYPIC DATA ON COMMERCIAL PIGS, INCLUDING WHOLE BODY COMPOSITION DATA FROM COMPUTER TOMOGRAPHY SCANS, DAILY BODY WEIGHT AND FEED INTAKE DATA, AND CARCASS DATA ON COMMERCIAL PUREBREDAND CROSSBRED PIGS. THIS PROJECT ADDRESSES THE 'NOVEL QUANTITATIVE GENETIC METHODS INCLUDING SELECTION THEORY AND MODELING' PRIORITY OF THE PROGRAM TOOLS AND RESOURCES - ANIMAL BREEDING, GENETICS AND GENOMICS PROGRAM AREA PRIORITY CODE - A1201 - BY INCORPORATING BIOLOGICAL KNOWLEDGE IN TO METHODS FOR GENETIC AND GENOMIC PREDICTION, WITH MAJOR ECONOMIC IMPACT ON LIVESTOCK PRODUCTION.