Kelci Miclaus
Advanced Analytics R&D Manager
JMP Life Sciences
Abstract:
Of the 7.3 billion people on this earth, 3 billion suffer from malnutrition. Feeding our growing population is one of the most demanding problems of our generation. How can we improve our crops to have higher yields, better disease-resistance, with ability to grow in harsher environments, while simultaneously increasing vitamin content? New prediction tools in JMP and JMP Genomics help breeders leverage plants’ genomic variability to find solutions to these complex multi-trait problems. In this presentation we show how JMP Genomics, a solution that combines JMP with SAS and extensive custom JSL applications, builds predictive models with genomic markers to optimize trait outcomes. Scoring code from models are used to make potential plant crosses and evaluate predicted outcomes of multiple traits simultaneously. Interactive JMP graphics using dominant row selection to select Pareto frontier points allow breeders to select potential plant crosses to make in the field that optimize multiple traits based on genetic inheritance. Essentially, with analytical prediction tools and interactive exploration, we can execute in-silico multi-year breeding programs in a matter of hours. Corn breeding data from the Brazilian Agricultural Research Corporation (Embrapa) will be used in demonstrations.