The UIC has been configured to provide a research platform to address scientific questions raised by Syngenta using a Systems Biology approach. On commencement of the UIC, two “pioneer” projects were chosen to initiate the collaboration; they focus on tomato fruit quality and predictive metabolism and toxicity. This was extended by a third project on ecosystems modelling, where we are applying the mathematics of the small scale (cells, genes and metabolites) to that of the large scale organisms and fields.
Tomato Ripening Project
The objective of this project is to build a predictive model of tomato ripening and fruit quality (colour, texture, flavour) which would allow Syngenta to identify the main genetic control points in this process.
Syngenta’s goal is to exploit this knowledge through conventional breeding programmes to develop tomato breeds with quality traits meeting both retailing requirements (e.g. tomato texture) and customer satisfaction (e.g. tomato taste).
The project aims to understand the whole process of fruit development and ripening. Although much of the specific biochemistry of fruit ripening is already known, a more holistic view of fruit ripening is being developed.
In this project, known genetic knockouts are used to look at changes affecting the process of fruit development and ripening. Data generated from the various analyses is then fed into the modelling work performed at the Imperial College. This project is focusing on tomato, but a similar approach could easily be applied to peppers and melons for example.
Predictive Toxicology Project
The objective of this project is to devise a predictive model describing key molecular events in xenobiotic induced-liver toxicity, which would in turn lead to the development of a toxicological model for use in Syngenta’s regulatory and research environments.
The desire is to apply Machine Learning approaches to multi-“omics” data to better predict the safety of lead chemicals before they go into expensive toxicology programs.
It is anticipated that, if successful, the model will be able to direct experimental design and choices to reduce the cost and number of experiments needed to predict the toxicological end points from a range of chemistries.
The overall aim is to identify mechanistic descriptions in model species, evaluate to what degree that can be applied to humans, and then apply the mechanistic insights gained to optimize the human safety of new and current Active Ingredients.
Ecological Modelling Project
Sustained increases in crop productivity may be impossible if the functions provided by biodiversity are diminished or eliminated by the direct or indirect effects of agricultural management. Ecological modelling has the potential to predict the effects of agricultural management on crop productivity and biodiversity.
However, current approaches to modelling have two important limitations: first, the models tend to be specific to particular crops, locations and taxonomic groups affected (single species or groups of closely related species); and secondly, the interaction between productivity and the functions provided by biodiversity is not considered.
A model that is applicable to any system of crop management at any site, and that can predict the joint effects of management on crop productivity and functional biodiversity, would therefore be extremely valuable.
The UIC Ecosystems Project seeks to ease the construction of food webs by using census data: surveys of the individuals that occur together at a certain place and time.
The basic premise of this work is that machine learning from co-occurrence data, along with background knowledge about the feeding relationships between different kinds of organisms, can automatically produce a description of the food web comprising the species, or functional groups of species, represented in the census data.
Network analysis offers a means to predict the effects of perturbation (e.g., agricultural management) on the functioning of food webs, allowing the development of decision-support systems for optimising the delivery of ecosystem services from agricultural land.