Nicole Cocks

Nicole CocksNicole Cocks

University of Wollongong

Biography

Nicole is currently an undergraduate in the School of Mathematics and Applied Statistics. She hopes to graduate from her Bachelor of Medical Mathematics degree in December 2014 at the University of Wollongong and complete her Honours year in 2015. She is interested in statistics and biology, which has been the focus of her previous studies and is the foundation for her current research project with AMSI. In her second year of study, she participated in a winter project with NIASRA to examine probabilities associated with leukemia clusters in NSW schools. At the end of her second year, she went on to another project with NIASRA, which investigated the most accurate method to classify an individual as pre-diabetic, overt diabetic or ‘healthy’. She looks forward to the completion of her studies at the University of Wollongong and is excited to move into a career in biostatistics.

Statistical Protocols for Late Maturity Alpha-Amylase in Wheat

The assessment of grain defect traits is assuming greater importance in wheat germplasm selection. Late maturity alpha-amylase (LMA) is one such characteristic that renders wheat unsuitable for highvalue end products, even though the grain may appear sound. The frequency of lines in wheatbreeding programs which LMA appears reasonably high, and hence LMA has become an importanttrait that is now routinely assessed in all wheat breeding programs in Australia.In contrast to grain yield, statistical design and analysis methods specifi c to the measurement ofLMA expression have been far less extensively researched. Like many grain quality traits in wheat,phenotyping for LMA involves a complex process involving experiments which are so-called multi-phase experiments. A multi-phase experiment is an experiment which involves several time periods,and has observational units which are completely diff erent to those from the preceding phases. Thesimplest multi-phase experiment involves 2 phases, and these were fi rst introduced by McIntyre(1955). A very common 2-phase experiment in agriculture involves a fi eld phase and a laboratoryphase. In the field phase, a fi eld trial is carried out on field (experimental) units and then the grainfrom that experiment is taken in to the laboratory where the experiment involves laboratory units.Phenotyping for LMA has at least 4 phases, most of which are unrandomised.In response to concerns with the accuracy of LMA expression experiments, Butler et al. (2009)developed a protocol to improve the design and analysis of LMA expression experiments. Thisprotocol accounts for non-genetic sources of variation in all phases by adoption of a model-basedapproach to design using the R package od (Butler, 2013). This protocol has been in use since2011 but there remain some unresolved issues. One of these issues concerns the relatively highresidual variation (i.e. relative to other sources of non-genetic and genetic variation). Preliminaryanalyses of the past four LMA expression experiments suggest that there may be ways of reducingthis variation by a more sensible compositing and laboratory randomisation scheme. The aim ofthis project is to examine this initial observation in more detail and consider the accuracy of variousalternative schemes which achieve higher accuracy, but are inexpensive and easy to implement.

References

Butler, D. (2013). On The Optimal Design of Experiments under the Linear Mixed Model. Ph. D.thesis, The University of Queensland.Butler, D. G., M. K. Tan, and B. R. Cullis (2009).

Improving the accuracy of selection for latematurity alpha-amylase in wheat using multi-phase designs. Crop and Pasture Science 60, 1202{1208.McIntyre, G. A. (1955). Design and analysis of two-phase experiments. Biometrics 11, 324{334.1

Contact Us

We're not around right now. But you can send us an email and we'll get back to you, asap.

Not readable? Change text.