Bionote
Agronomist, lifelong curious, passionate about agriculture, plant breeding, data science, and statistics. I developed my interest in plant breeding as an intern at a Dutch seeds company (Nickerson Zwaan B.V.). After graduation, I worked as a sales specialist and consultant for 5+ years in irrigation engineering. Then, I decided to come back to academia and deepen my knowledge of plant breeding and genetics.
Currently, I am working with plant breeding, quantitave genetics, molecular biology, bioinformatics, and statistics, using genomic and breeding tools to enhance plant architecture in tomato crops by utilizing the brachytic gene present in short internode plants.
Presentation Abstract
Ideal plant architecture (PA) is a key challenge in modern agriculture, as rapid changes in crop production require cultivars to adapt quickly to meet farmer and consumer demands. Modifying tomato PA can improve the efficiency of open-field mechanization, urban farming, and home gardening. Although dwarf genotypes have been used successfully in tomato and other crops, the available genes often impact other important traits such as fruit size, flowering time, and stress resistance. Generation mean analysis (GMA) is a valuable tool for assessing genetic variances and gene interactions across traits, but it has been largely overlooked in favor of more advanced statistical methods. In this study GMA is applied to six generations of a biparental cross of contrasting tomato plants—dwarf and tall varieties—we combine spatial analysis and generalized linear mixed models (GLMM) in a two-stage framework to better understand the genetic architecture of plant architecture traits. Our results showed that irrigation had a minor effect compared to plant positioning. The two-stage approach reduced the relative contribution of the mean in both additive-dominance and epistatic models, highlighting stronger signals of gene interactions. This allowed us to correctly assign genetic models to five traits previously mislabeled. Further work is needed to incorporate additional environmental covariates and a wider range of traits, which also opens opportunities for applying alternative statistical approaches, such as multi-trait. Here we also propose the underlying genetic architecture for the fifteen PA traits evaluated in this study.
Co-Author (Supervisor): Alex Lipka