b'have understanding basic concepts and data analysisRadical Collabsprocesses.As researchers like Jarquin become increasingly able Oftentimes students pay for their data analysis toto collect new data, new collaborations are needed to be done and even when this is carried out, there is noensure that data are effectively used and identify gaps guarantee that the concept or the software output isthat need closing.understood or if it was analyzed correctly, she says.Radical collaborations were essential to advancing The greatest impact is that African students are gain- DNA sequencing technology, to where it now costs ing confidence in analyzing and interpreting their dataabout $100 to re-sequence the human genome. This and most importantly they teach others in their varioushas resulted in a deluge of low-cost genomic data.schools and communities. In contrast, phenotyping is a 10,000-year-old chal- Reaching as Discovering the field of data analysis and statis- lenge that started with the visual assessment of plantsmany people as tics was what led Diego Jarquin to make a name forfor desirable attributes at the time of plant domestica- possible in himself as one of the top statisticians in his field. Borntion. Phenomics technologies have only just begun toorder to in Oaxaca, Jarquin received a PhD in statistics fromevolve more rapidly, and phenomics data are typicallytransform the the University of Postgraduate Education in Mexico inmore expensive to collect on a large scale than theirdiscipline of 2012. He had postdoctoral training at the Universitygenomic counterparts. crop of Alabama-Birmingham and at the University ofThis has caused a massive imbalance in the volumeimprovement is NebraskaLincoln. He has been at Nebraska since 2017.of data collected, says Michael Gore, Liberty Hydea mission of Bailey Professor and professor of molecular breedingBorlaug Scholar and genetics at Cornell University. Gore spoke at theCatherine NAPB virtual meeting as part of a session on high- Danmaigona throughput phenotyping and the promise it holds forClement.bolstering crop production.The concept of field-based high throughput phe-Radical collaborations among plantnotyping has been advocated by those like Gore in the plant breeding community to reduce human effort, biologists, engineers and computer scientistscollect higher amounts of data, evaluate large popula-are needed to bring the lab to the field.tion sizes at many time points in a growing season, decrease subjectivity when scoring plants and reduceMichael Gore phenotyping costs over time.Radical collaborations among plant biologists, engineers and computer scientists are needed to bring the lab to the field, allowing multi-scale data to be col-lected, he says. Statistician Gores team began collaborating in 2010 withDiego Jarquin Jarquin was awarded the 2020 NAPB Early Careeragricultural engineers at the University of Arizona,was awarded Scientist Award and spoke at the NAPBs virtual meet- implementing a rapid phenotyping platform to monitorthe 2020 NAPB ing about how working with CIMMYT as a PhD studentchanges in physiological phenotypes at multiple timeEarly Career and coming to the United States allowed him to turnpoints within a day and multiple days across the grow- Scientist Award. his passion for data science into a career.ing season in cotton.PHOTO: UNIVERSITY OF In the kind of studies that we perform, every dayThe goal, as explained on the Gore lab website, isNEBRASKALINCOLNthere are new layers of information that were not thereto understand the pattern of QTL expression dynam-before, he says. ics under varying environmental conditions and more When I had a chance to work with more layersspecifically, what QTL may be important in mitigating of information, that was when I realized that to makethe consequences of heat and drought stresses.sense of all the different available omics of information,Shifting focus from the ground to the sky, Gores something different needed to be done, like developingteam has assessed the potential use of unoccupied new models and methods able to distinguish the dif- aerial vehicles (UAVs) as efficient co-workers for pre-ferent data types during analyses such that we can getcisely monitoring the incidence and severity of north-the most of the information out of the data. ern leaf blight (NLB) disease in maize. WHERE Jarquin has worked with plant breeders of differentThrough a collaboration with Hod Lipson atON THE WEBcrops in the U.S., Mexico, Japan, Australia, and fromColumbia University and Rebecca Nelson of Cornell many other countries. He says the biggest lesson hesUniversity, Gores team is using aerial imagery collectedFor comprehensive learned is that no matter the place on the earth, all thevia small UAVs in combination with deep learning tocoverage of this years NAPB virtual meeting, researchers with whom I had the privilege of workingrecognize leaf blight symptoms in plant images. Theincluding podcast with are interested in developing products, methodsultimate goal is to develop a semi-autonomous systeminterviews and award and methodologies to help humanity fight hunger.that can identify and quantify levels of NLB and otherwinner profiles, visit Their passion touched me and every single day. potential plant diseases in the field.SW seedworld.com/napb-202186/ SEEDWORLD.COMINTERNATIONAL EDITION 2022'