Dario Grattapaglia - EMBRAPA

Dario Grattapaglia

1990 ‐ 1994 Ph.D. Genetics and Forestry (co‐major) North Carolina State University, Raleigh, NC, USA ‐ Advisor: Prof. Ronald Sederoff

Elected to the Honor Society Phi Kappa Phi, Chapter 33, 1992
1981 ‐ 1985 B.S. Forest Engineering, University of Brasilia, Brazil 

1994 ‐ present Research Scientist ‐ Project Leader, Plant Genetics Laboratory EMBRAPA – Genetic Resources and Biotechnology

2000 ‐ present Professor – Graduate Program in Genomic Sciences and Biotechnology Catholic University of Brasilia
2016‐ present Adjunct Professor ‐ Department of Forestry and Environmental Resources
North Carolina State University, USA
1996 ‐ present Co‐founder and director of Heréditas ‐ Tecnologia em Analise de DNA Private DNA analysis laboratory for human plant and animal forensics and plant molecular breeding
1995 ‐ present Associate Faculty at University of Brasilia ‐ Dept. of Molecular Biology
1994 Assistant Professor at UNESP State University of São Paulo ‐ Dep. of Genetics
1985 ‐ 1990 Research Scientist – Dept. of Plant Cell Biology ‐ Bioplanta Tecnologia de Plantas Ltda. (Joint venture between Native Plant Inc. ‐ South Lake City and BAT ‐ British American Tobacco) ‐ Campinas, São Paulo, Brazil


The convergence of genomics and quantitative genetics: genome-wide prediction of complex traits in forest trees

Planted forests supply woody biomass in a sustainable fashion that would otherwise come from deforestation and forest degradation. They also provide environmental services in erosion and water cycle regulation and act as long-term carbon sinks. The challenge of tree improvement programs is, however, the long interval between the breeding investment and the deployment of improved material. After twenty-five years of forest tree genomics research, and despite important advances in QTL mapping and association genetics (AG), genomics has not had any significant impact in operational tree breeding. Reasons include the limitations of early genomic technologies, the genetic heterogeneity of largely undomesticated trees and, mainly, the overly optimistic and rather naïve outlook about the architecture of complex traits. The advent of high throughput genotyping technologies coupled to genomic selection (GS) have provided a new paradigm to integrate genomics and quantitative genetics into breeding. By fitting thousands of genome-wide markers concurrently in predictive models, GS can capture most of the ‘missing heritability’ of complex traits that QTL and AG classically fail to explain. The milestone publication of the Eucalyptus grandis genome allowed us to develop a multi-species SNP genotyping chip based on whole-genome resequencing of 240 Eucalyptus tree genomes of 12 species. Using this genotyping platform in a number of industrial breeding programs in Brazil, we have shown that GS accuracies can match or surpass conventional phenotypic selection for growth and wood properties traits. GS significantly reduces the length of a breeding cycle by applying ultra-early selection of genomically multi-trait ranked seedlings, precluding the progeny trial stage. Top ranked seedlings can be subject to early flower induction and inter-mated to create the next breeding generation and/or immediately propagated and deployed as clones in validation field trials. We have found, however, that GS predicts poorly across unrelated populations and variable environments, therefore requiring breeding-population-specific GS models. Genome-wide prediction brings a new perspective to the understanding of quantitative trait variation in forest trees and shall make genomics finally find its way into applied breeding. Strategic and logistics aspects of operational GS adoption are now the challenges faced for its full integration into routine tree breeding operation.