Data Overload—Breeding Decision–Support Software to the Rescue!

Wednesday, August 1, 2012: 9:15 AM
Trade Room
S. Jung , Washington State University, Pullman, WA
Taein Lee , Washington State University, Pullman, WA
Kate Evans , Washington State University, TFREC, Wenatchee, WA
Cameron Peace , Washington State University, Pullman, WA
Gennaro Fazio , USDA–ARS, Geneva, NY
Sushan Ru , Washington State University, Pullman, WA
Amy F. Iezzoni , Michigan State University, East Lansing, MI
Doreen Main , Washington State University, Pullman, WA
Crop improvement programs have become increasingly sophisticated with the advent of DNA-informed breeding. Now breeders not only have to keep track of the pedigree records and phenotypic data for their selections, but  also genotypic data and its meaning. Utilizing this vast amount of data to make the best crossing and seedling selection decisions can quickly become overwhelming. In RosBREED we are addressing this challenge with development of an online Breeding Information Management System (BIMS) to organize and handle breeding data in a systematic manner to support breeding decisions. BIMS provides interfaces where breeders can search their data by dataset, location, variety, trait with threshold limits, marker allele, and pedigree. From the results page, breeders can download phenotypic and genotypic data as well as view detailed data of individual varieties. BIMS resides within the Genome Database for Rosaceae (GDR), and the breeding data are linked to the related genetics and genomics data for further investigation. BIMS also provides a tool to choose a variety, its progenitors and descendants in the pedigree, and phenotypic and genotypic data of this pedigree to produce an input file for the publicly available software, Pedimap, which enables the visualization of these data. Also available in BIMS are two tools that support breeding decisions.  The “Cross Assist” module, already available, with new functionalities being added as suggested by users, helps identify efficient parental combinations that provide a high proportion of seedlings with desired performance levels. The “Seedling Select” module, still in prototype form, helps breeders identify efficient seedling selection schemes by integrating DNA tests into routine breeding operations with early-stage culling of seedlings that do not contain desired performance levels and retention of those that do. Both these tools help breeders generate better progenies with available resources—saving thousands of dollars otherwise spent evaluating inferior progeny.