GENOMIC SELECTION HAS BEEN REVOLUTIONARY FOR ANIMAL BREEDING, BUT CURRENT GENOMIC BREEDING VALUES DO NOT ACCOUNT FOR ENVIRONMENTAL DIFFERENCES ACROSS PRODUCTION ENVIRONMENTS. THERE IS A NEED TO QUANTIFY THE EFFECTS OF CLIMATE VARIABLES ON LIVESTOCK PRODUCTION, UNDERSTAND HOW THEY INTERACT WITH EACH OTHER AND WITH THE GENETIC COMPONENTS OF IMPORTANT TRAITS. THE MAIN OBJECTIVE OF THIS PROJECT IS TO DEVELOP PREDICTION MODELS THAT MAKE USE OF LARGE-SCALE ENVIRONMENTAL AND GENOMIC DATA TO MODEL GENETIC, ENVIRONMENTAL AND GENE-BY-ENVIRONMENT INTERACTIONS. IN THIS PROJECT WE WILL DEVELOP NOVEL STATISTICAL AND COMPUTATIONAL APPROACHES FOR PREDICTION METHODOLOGY TO BE ABLE TO ADDRESS THE DIMENSIONALITY PROBLEM CAUSED BY THE CONTINUOUS NATURE OF CLIMATE VARIABLES AND THEIR MULTITUDE OF POSSIBLE PARAMETERS AND INTERACTIONS. WE WILL USE TIME SERIES ANALYSIS AND ARTIFICIAL INTELLIGENCE METHODS TO OPTIMIZE STATISTICAL PREDICTION MODELS THROUGH FEATURE SET AND MODEL SELECTION. EXPECTED OUTCOMES FROMTHIS WORK ARE AN ENVIRONMENTALLY INFORMED PREDICTION METHODOLOGY TO ASSIST IN FOOD PRODUCTION AND NEW INSIGHTS INTO THE EFFECTS OF CLIMATE ON LIVESTOCK PRODUCTION. PREDICTION MODELS THAT ACCOUNT FOR ENVIRONMENTAL PARAMETERS WILL BE A POWERFUL MANAGEMENT TOOL FOR AMERICAN AGRICULTURE AND SHOULD HELP GUIDE SELECTION DECISIONS IN DIFFERENT ENVIRONMENTS. WE EXPECT THAT NEW RESULTS CAN BE IMMEDIATELY DELIVERED TO THE US CATTLE INDUSTRY - OUR PRIMARY CASE STUDY - TO IMPROVE PREDICTION ACCURACY AND INFORM BREEDING SCHEMES. QUANTIFICATION OF THE EFFECTS OF CLIMATE ON PRODUCTION COULD BE BROADLY SIGNIFICANT FOR THE DISCUSSION ON WHICH IS THE BEST PATH FORWARD TO SELECT FOR IMPROVED PRODUCTION AND LONG-TERM RESILIENCE IN A CHANGING ENVIRONMENT.