Predicting bioactive molecules for plant health and agricultural applications

A platform for functional prediction and validation of undiscovered secondary metabolites in Actinobacteria.

Project duration: January 2023 – June 2026

Image: Shutterstock

Dr Marta Gallart Diumenge

Research Scientist


Dr Louise Thatcher, Dr Rosangela Devilla, Dr Oliver Mead, Dr Robert Speight

The challenge

Traditional pesticide use is declining globally due to public concerns, market access and regulatory constraints. Biopesticides, such as microbial-derived bioactive molecules, offer a safer, sustainable and more effective solution to maintain high productivity agricultural systems. Biopesticides have a number of potential benefits such as lower ecological impact, and increased effectiveness against pests, weeds or pathogens that have developed resistance towards traditional synthetic pesticides.

Members of the Actinobacteria phylum are a rich source of bioactive natural molecules, including herbicidal secondary metabolites. Advances in next generation sequencing and analysis allow us to predict secondary metabolite gene clusters and indicate that Actinobacteria can produce many more secondary metabolites than previously thought. The challenge is how can we predict the chemistries and the biological function of these novel natural molecules to fast-track the development of new biopesticides, and how can we produce them cost effectively at scale for commercial uptake?

Our response

Using integrative omics, engineering biology and metabolic modelling technologies, we are helping to solve these challenges by developing a high-performance pipeline for the prediction and discovery of novel bioactive molecules encoded by secondary metabolite gene clusters in Actinobacteria.

The project aims to exploit a CSIRO Actinobacteria strain collection to uncover its hidden chemical potential and create a database that integrates genomes, advanced genome mining, phenotyping and metabolomics data, followed by downstream chemical and biological characterisation. Low-cost commercial-scale production will be achieved through domestication and engineering of our non-model Actinobacteria strains or the transfer of gene clusters to production organisms.

Using industry needs to guide project design, we will use herbicidal activity as a case study to identify Actinobacteria bioactive molecules with phytotoxic activity, otherwise known as bioherbicides. This research will be a paradigm change in the identification of new bioherbicides and the development of tools and models that enable agriculture and society to benefit from non-model organisms with unique traits.


The search for new bioactive molecules is hindered by our ability find them, to predict their function and biological activity, and to produce them at low cost. This project aims to overcome these obstacles through new, faster, and more accurate predictive methods coupled with high throughput, miniaturised experimental systems for functional testing.

The Actinobacteria-based bioactive molecule discovery, functional prediction and precision bioproduction pipeline will advance us beyond traditional ‘one strain one molecule’ approaches to a foundational platform using streamlined approaches for fast-tracked bioproduction of multiple molecules. Such potential can be tailored towards other crop protection products (biofungicides, bioinsecticides), or leveraged towards pharmaceuticals, other agrichemicals, or the production of food additives.

More information

Innovative bioproduction platforms