As climatic conditions change, farming staple crops such as maize, barley, sorghum and tomatoes becomes significantly more challenging.

Meanwhile, bacteria can assist the plants to become more resilient to changing heat, drought and flooding – in a cost-efficient way. The key is to identify these helpful bacteria quickly enough so that the agricultural industry can create reliable products for farmers.

At the UJ Research Centre for Plant Metabolomics, scientists are steadily contributing to a global research collaboration using Artificial Intelligence to do exactly this.

A search tool with data from researchers around the world is tracking down microbes and their metabolites in a game-changing new way. Uniquely, it can significantly speed up the identification of known and unknown bacteria, fungi and single-celled archaea, and their close relatives.

The tool can be used to improve the resilience of food crop production amidst a changing climate; to analyse the human gut microbiome; and many other applications.

UJ researchers are contributing ‘fingerprinting and ancestry’ microbial metabolomic data to the project, says Dr Fidele Tugizimana. They are also testing the computational tool and workflows around it.

Tugizimana is the incoming Co-Director for the UJ Research Centre for Plant  Metabolomics. The Centre has published several research papers on bacterial metabolites that can improve the resilience of tomatoes, sorghum, maize (corn) and barley.

The global GNPS microbeMASST project is led by Prof Pieter C. Dorrestein at the University of California San Diego (UCSD). Curated data from 25 research teams, including the UJ Research Centre for Plant Metabolomics, containing more than 60,000 microbial monocultures, forms the core of the search tool.

In a recent journal article in Nature Microbiology, Dorrestein and UCSD co-authors Dr Simone Zuffa, Dr Robin Schmid, and Dr Anelize Bauermeister introduce microbeMASST as part of a wider ecosystem of computational tools, facilities and equipment called GNPS.

The unique strength of microbeMASST is finding the possible microbes that can produce a particular metabolite within its AI-powered ‘search-engine’. The ‘search-engine’ is built from the metabolomic data contributed by the global research groups.

In laboratory and greenhouse tests, researchers may find a ‘soup’ of beneficial microbes that appear to help crop plants in some ways, says Tugizimana. But there is a lot to discover before the ‘soup’ can be ‘packaged’ in a liquid or carrier-material formulation and sold as an agricultural product.

Farmers are dealing with enough uncertainty already. The products they buy need to have a reliable and predictable effect on their crops.

“You need to know what type of chemistry those microbes in the formulation produce,” adds Tugizimana.

“I should be able to tell you that the formulation will help your specific crop resist a harsh environment, to produce a higher yield, and why that product would have that effect.”

The growing mountain of microbe metabolite data in microbeMASST is designed to be searchable in a specific way, explains Dorrestein via email.

“To make data ready for data science, including AI applications, is the key goal of microbeMASST. It was curated with around 100 scientists worldwide by taking metabolomic data from microbes grown in culture and curating the taxonomic information.

Crops that could reliably be grown in the past are likely to have lower yields in future, due to global warming. Here the GNPS microbeMASST project can contribute to a more resilient future agriculture.

“Climate change is a certainty,” says Prof Ian Dubery, from UJ Biochemistry. Dubery is the outgoing Director of the UJ Research Centre for Plant Metabolomics.

“What we know is that traditional phenotypes, for example, disease resistance, are breaking down during secondary stress [created by climate change]. The combination of stresses – heat, drought, pathogens – is certain to have detrimental effects on agricultural production, not only in South Africa but globally.

“We need to look at alternative approaches to agricultural practices. MicrobeMASST can certainly assist in enhancing future research globally. The tool can help speed up development of new microbial ‘cocktails’ or consortia to apply to seeds, soil or plants, to improve yield or resilience,” concludes Dubery.

This article was originally published in the September 2024 issue of the UJ Research and Innovation magazine at https://www.uj.ac.za/research-at-uj/uj-research-and-innovation-magazine/