Page - 121 - in Intelligent Environments 2019 - Workshop Proceedings of the 15th International Conference on Intelligent Environments

PredictiveCharacterizationof ICARDA GenebankBarleyAccessions singFIGS andMachineLearning ZainabAZOUGHa,b,ZakariaKEHELb,AzizaBENOMARa,MostafaBELLAFKIHa andAhmedAMRIb aINPT,Rabat,Morocco bICARDA,Rabat,Morocco Abstract. The International Center for Agricultural Research in the Dry Areas (ICARDA)has auniquegermplasmcollectionof barley, amongmanyother crops that it holds in its genebank. This collection contains landraces and barley wild relatives andmost of themare georeferenced.Distribution of genetic resources is a core genebank activity aiming at responding to requests from various users in- cluding breeders, researchers, farmers, etc. ICARDAhas developed over the last decade an efficient approach for better targeting adaptive traits called theFocused Identification ofGermplasmStrategy (FIGS). FIGSapproach links adaptive traits toenvironments (andassociatedselectionpressures) throughfilteringandmachine learning and it focuses on accessions that aremost likely to possess trait specific geneticvariation. In thispaper,wepresentaworkofpredictivecharacterizationon ICARDAbarleycollectionusing theFIGSapproachand its algorithmscombining severalmachine learningmethods, and using several characterization traits.Most of the studied traits have shownahighpredictability.Outcomes from this analysis are then used tomake a predictive characterization of the entire ICARDAbarley collectionbyassigningprobabilitiesof each trait to thenon-evaluatedaccessions. Keywords.barley, characterization,FIGS,machine learning 1. Introduction Genebanksworldwidehold collections of plant genetic resources for long-termconser- vation andmaintain crop diversity for current and future use by crop improvement re- search,directuseand training.Mostofgenebanksare facingmajorproblemsofsizeand organization. Some collections have grown so largemaking theirmain activitieswhich aretheconservationandtheuseofthegeneticdiversitychallenging.Anotherchallenging aspectofplantgeneticresourcesconservationis the lackof informationaboutaccessions specificallyapreciseevaluation information.This ismainlydue to thechallengeofeval- uationtheentirecollectionofagenebank.However, severalgenebankshavedoneagreat jobonmaintainingandcuratingpassport information.Butpassport datadoesnothelpa user of the genebank to discernwhich accession in a database is potentially containing the trait of interest. The concept of core andmini-core collections have been proposed as a strategy that allows theuseof small portionof agermplasmcollection to represent U Intelligent Environments 2019 A. Muñoz et al. (Eds.) © 2019 The authors and IOS Press. This article is published online with Open Access by IOS Press and distributed under the terms of the Creative Commons Attribution Non-Commercial License 4.0 (CC BY-NC 4.0). doi:10.3233/AISE190031 121