Stats for Agriculture: How to deal with Big Data?
We recently hosted a hybrid event with the James Hutton Institute at their Dundee site entitled “Stats for Agriculture: How to deal with Big Data?“. We were delighted to have Brian Cullis and Alison Smith from the University of Wollongong visiting us, as well as legend in the field Rosemary Bailey.
The workshop started off with Rosemary giving a convincing talk on why agricultural and environmental research still needs statisticians, with examples from her long career in the design of experiments.
We then heard from Brian and Alison, both of whom have long-standing interests in rigorous and pragmatic design and analysis of large, complex field trials and the demystification of the process. Their part of the workshop focused on dealing with analysis of the increasingly large and complex data sets we are beginning to generate, particularly with the development of new sequencing, drone, robotics, continuous time monitoring and high throughput phenotyping technologies. Brian gave a superb talk on model-based design of selection experiments, followed by Alison speaking on selection in plant breeding experiments. Brian then put the theory into practice with a quick overview of their R package DWReml - software for fitting LMMS in a fast and easy R environment to any sort of plant breeding data.
Nick Schurch from BioSS then gave a thought-provoking talk about challenges and opportunities in this field in the context of Open Science. Then followed wider discussion about current challenges, solutions and the potential for future collaborations.
Many thanks to Brian, Alison and Rosemary for their contributions to this workshop!
If you missed the workshop, you can watch a recording of the event here.