Computers, Roots & Big Data from the Field: Can new methods identify uncharacterized phenomena in existing data?
Speaker: Dr. Alexander Bucksch (University of Georgia)
watch the recorded webinar online
Abstract:
The use of mathematical and computational methods in the plant sciences enables two new major areas: The discovery of uncharacterized phenomena in existing phenotyping data and detailing phenotypic measurements beyond the manually achievable. Therefore, plant phenotyping is able to set new breeding targets for rapid increases of grain yields or improved plant functions such as carbon sequestration to lower atmospheric CO2 concentrations. However, the opportunities to phenotype discovery are buried in large amounts of collected phenotyping data that often is only analyzed for one experimental purpose. The computational plant science lab at the University of Georgia makes use of the increased data volume and physical resolution of imaging data to quantify root architecture. In particular, the coupling of high-performance computing with image acquisition systems holds the potential to discover uncharacterized phenotypes that were previously “invisible” to phenotyping researcher. In part, the invisible phenotypes can be attributed to the time consuming manual evaluation of root phenotypes that resulted in a low coverage of the phenotypic variation. This IPPN webinar will highlight the mathematical challenges in phenotyping root architecture associated with the low coverage phenotypic variation. As a response to these challenges, we introduce computational advances developed at UGA to quantify variation root phenotypes on the population level. Furthermore, the webinar will give a preview of 3D imaging and high-performance computing technologies that will further detail the observed variation of phenotypes from the field.
CV:
Alexander Bucksch received his PhD from the Delft University of Technology in the Netherlands in 2011, where he developed the first algorithm to measure complete tree crowns from laser scanned trees. He then moved as a PostDoc to the Georgia Institute of Technology where he was jointly appointed between the School of Biology and the School of Interactive Computing and began to work on root phenotyping. In 2016, Dr. Bucksch joined the faculty of UGA as an Assistant Professor with joint appointment in the Department of Plant Biology, the Warnell School of Forestry and Natural Resources and the Institute of Bioinformatics. He leads the computational plant science lab which combines computer vision and shape analysis to analyze the topological and geometrical characteristics of plant architecture. The lab is best known for DIRT (Digital Imaging of Root Traits, http://dirt.cyverse.org), which is the world’s largest root phenotyping platform with over 620 users in over 40 countries. Dr. Bucksch is a founding member of the Georgia Informatics Institutes and the Plant Phenomics and Robotics Center at the University of Georgia. He received the 2020 Early Career Award from the North American Plant Phenotyping Network and the NSF CAREER award in 2019 to quantify and simulate the phenotypic spectrum of roots that he presents today.