The Data Catalyst
Goals
This team aims to develop a ‘data catalyst’. This novel approach facilitates studying plant resilience when a plant is exposed to combinatorial stresses. It is based on the integration of experimental data with mechanistic modelling, machine learning and bioinformatics. The goal is to develop the data catalyst approach in a generalizable, modular and robust manner. In this way, the data catalyst can be extended and deployed in other work packages of CropXR.
Stress matrix experiments with thale cress seedlings. The left 3 seedlings are grown in a well-watered soil and the right in a dry soil.
Approach
The research focuses on the responses of thale cress or Arabidopsis thaliana seedlings to combined and isolated exposure to drought and elevated temperature. To obtain insights, the team collects (multi-omic) data from stress matrix experiments. Furthermore, it develops models that can reproduce and predict seedling responses to temperature and drought stress. The team develops both machine learning models and mechanistic models. These models are integrated into each other and validated in an experimental way by an iterative modelling-experimentation cycle.
Activities
In the upcoming time, the team will finalize various protocols related to data collection, methodologies and (performance) measurements. Moreover, methods for more automated phenotyping will be developed. The team will extend mechanistic modelling modules with additional parameters. Methods will be developed and refined related to networks and driver genes to enhance plant architecture modelling. Lastly, the team works on control and sensitivity analysis methodology.
Illustration: overview of the interplay between experiments, data analysis, machine learning and modeling to arrive at integrated model for plant stress responses.
Team
Work Package leader Kirsten ten Tusscher, Professor Computational Biology, UU
Aalt-Jan van Dijk, Professor Data Analysis, UvA
Alex de los Santos Subirats, PhD candidate, TU Delft
Ben Noordijk, PhD candidate, WUR
Bianca Cosma, PhD candidate, TU Delft
Christa Testerink, Professor of Plant Physiology, WUR
Dick de Ridder, Professor Bioinformatics and AI, WUR
Francesca Giaume, Postdoc, WUR
Jelle Keijzer, PhD candidate, UU
Jorn de Haan, PI, Genetwister – industrial partner
Kimm van Hulzen, PI, Genetwister – industrial partner
Laurens Krah, PI, KeyGene – industrial partner
Luca Laurenti, Assistant Professor, TU Delft
Mahshad Keshtiarast Esfahani, PhD candidate, TU Delft
Manuel Mazo Espinosa, Associate Professor, TU Delft
Marcel Reinders, Professor Pattern Recognition and Bioinformatics, TU Delft
Marcel van Verk, Research Director, Keygene – industrial partner
Martijn van Zanten, Associate Professor, UU
Milan van Hoek, PI, KeyGene – industrial partner
Monica Garcia Gomez, Assistant Professor, UU
Nikkie van Bers, PI, Genetwister – industrial partner
Pinelopi Kokkinopoulou, Technician, WUR
Thea van den Berg, PI, KeyGene – industrial partner
Thijs van Loo, PhD candidate, UU