Kansas State University, Langston University and Oklahoma State University have NSF/EPSCoR funding to combine simulation models, high-throughput plant/soil sensing, and genetics to predict wheat traits. Doing so is mandatory for annual rates of yield gain to meet global food needs at 2050. To contribute to project workforce development objectives and to overcome the difficulty of recruiting graduate students for multidisciplinary research, three project faculty (from Mathematics, Statistics, and Agronomy) developed a Problems course for graduate students and advanced undergraduates to (1) enhance a wheat computer simulation model and (2) evaluate an emergent statistical algorithm for predicting traits from genomic data. (The project aims to merge these approaches.)

The presentation covers the resulting “Math 799”, which was taught for two semesters. Included are: recruitment methods (word-of-mouth to hanging posters around campus), class demographics, topics taught and results achieved through student research (differential equations to hierarchical Bayesian models), an external evaluation of the effort, and student outcomes (including five accepted abstracts and professional presentations). Downsides are also discussed including a final report delayed for nearly a year by sundry communication complications. That problem’s resolution is presented along with Do’s/Don’ts to help others implement this very effective research/education method.

Presenter(s)

Stephen Welch