Unlocking the soil microbiome

How are data science techniques helping us better understand the microbial universe of the soil? Anthony Finbow of Eagle Genomics – a firm working at the intersection of biology, data science and bioinformatics – explains, and relates how this knowledge provides a more solid underpinning for sustainable agricultural innovation.

Over-reliance on nitrogen fertilizers is leading to the collapse of soil biodiversity. Traditional agricultural practices use fertilizer to optimise the nitrogen available to plants, bringing undeniable benefits ranging from higher crop yields to reduced production costs, consistent harvests to cheaper food. Yet, some predictions warn that there are only 60 harvests left in the world’s soil because it is so depleted by nitrogen fertilizers.

Soil plays a major part in the carbon capture cycle. Trees capture carbon monoxide, then transmit the carbon through to microbes in the soil, where it is held. The collapse of soil biodiversity is happening at microscopic scale. The nano ecosystem in the soil is facing a global threat, yet we know worryingly little about it ¬only 1% of the microbial universe of the soil has been examined and categorised in detail.

To reverse the effects of the climate crisis we must focus now on helper microbial communities found in soil. If we could replenish the source microbial universe, we will boost the ground’s ability to capture and retain carbon, and nature can reverse climate change for us—perhaps in as short a time as 30 years. This can only happen when we have a much better scientific understanding of the soil nano ecosystem.

Microbes are also the fundamental building blocks of the food chain. Supporting the soil microbiome will not only reverse climate change, but also help deliver the food that will be needed for the global population of 10 billion people forecast by 2050.

The microbiome holds the key to soil’s carbon capture. The interactions and effects of different combinations of microbes are well illustrated in the recent documentary, Kiss the Ground, narrated and featuring Hollywood star Woody Harrelson. This ground-breaking documentary reveals how, by regenerating the world’s soils, we can rapidly stabilise the earth’s climate, restore ecosystems and secure ongoing abundant food supplies. The filmmakers clearly demonstrate soil’s central role in capturing atmospheric carbon.

Advanced data techniques

Sophisticated technology applied to a wide range of scientific data is beginning to play a vital part in helping us understand these processes. Modern data management, applying network science using graph data structures and AI-based analytics, is allowing everyone from climate change scientists, soil experts and agri-manufacturers to benefit from microbiome-based knowledge analytics and discovery.

Advanced data techniques include fit-for-purpose data fabrics, multi-layered hypergraphs, AI-enhanced graph modeling and causal inference programming techniques. We will take a look at each of these.

Fit-for-purpose data fabrics frameworks ensure a sufficiently comprehensive set of data is available and accessible, has integrity and value, and can be exchanged, compared and understood in reliable and meaningful ways by non-data scientists.

Multi-layered hypergraphs are based on graph database technology that provides insight into multi-dimensional interrelationships between data, especially in complex scientific projects. The next step up is a hypergraph, or a multi-layering of multiple graphs. Stacking graphs allows scientists to gain greater microbiome insight though understanding more about the connections between the data. A standard graph representation reveals a data relationship across two nodes, linking an e.coli bacteria with a chemical reaction to fix nitrogen for example, while a hypergraph would explore multi-dimensional relationships.

Artificial intelligence (AI)-enhanced graph modelling can help decode patterns that might not be obvious to the human observer. AI lets biologists and other scientists uncover findings without needing to call on help from data scientists to manipulate the data or experts in statistics to decipher connections.

Causal inference programming techniques help scientists understand causality between data points. Advanced application of causal inference programming distils and investigates associations between diverse data, allowing scientists to delve deeper into root cause analysis.

 

Data-driven solutions

Scientists can take the data insights and apply practical solutions. Pioneering organisations are beginning to act on advances in understanding. For example, one innovation centres on engineering microbes to fix more nitrogen to plant roots so those plants can grow more effectively, avoiding the damaging environmental impacts of nitrogen fertilizer products.

Plant microbiomes are also being applied to increase yields and improving salt and drought tolerance of crops[1]. Soil microbiomes can be applied as bio-fertilizers for soils and can reduce nitrogen leaching[2]. Practical alternatives to non-biodegradable plastics that are the scourge of the oceans are under development. Amongst other applications, my company is developing technology to support enterprises working to develop hard-printed circuit-board materials which have been manufactured through bioprocesses exploiting microbiome science, to remain stable through their life, but which also degrade rapidly and appropriately after their useful life.

Microsoft has recently gone on record to predict that the food industry “has not even scratched the surface when it comes to understanding the microbiome”—but that once we have, there is a wave of innovation coming from microbiome science as more is discovered about the interplay of microbes and foods, its impact on human and animal health, as well as establishing what specific soil conditions will drive crop yield and improve animal farming.

Humanity’s biggest challenge

Addressing climate change while securing the future of humanity is our biggest challenge. Understanding the soil microbiome is crucial. Bill Gates maintains that understanding the microbiome will be “as big a breakthrough as anything else we will do over the next two decades.” Climate change scientists, soil experts and agri-manufacturers need to understand the role of micro-organisms and their interactions both in relation to each other and with the active ingredients of agri-chemicals. This relies on the ability to be able to combine and analyse huge volumes of biological data and reliably interpret any correlations and insights generated.

There are still some practical hurdles to overcome, especially around data standardisation, but innovative data technologies are already working with massive quantities of complex, multi-dimensional data to power discoveries about the soil microbiome. This growing understanding will drive the development of more sustainable agricultural practices leading to greater food security and addressing the urgent issue of climate change.

  • The author is CEO at Eagle Genomics, a UK firm applying network science to biology linked to the microbiome. It is working with 5 of the top 10 household and personal care companies in the world to create and launch new products that work in harmony with the human and environmental microbiome

Could a food-as-software model prove critical?

In 2019 independent think tank RethinkX published a widely-read report that introduced a ‘food-as-software’ model, which, it predicts, will supersede our current industrialised, animal-agriculture system. In the food-as-software model, foods are engineered by scientists at a molecular level and uploaded to databases that can be accessed by food designers anywhere in the world.

This could result in a more distributed, localised food-production system that is more stable and resilient than the one it replaces, stated the study. It notes how by 2035 US demand for beef and dairy products could be down by nearly 90%, leaving only local specialty farms in operation, while industrial farmland values may collapse by 40-80%. Lands previously used to produce animal foods could become a major carbon sink leading to environmental sustainability and renewal.

The food-as-software model relies on precision fermentation, a process that enables the programming of microbial organisms to produce almost any complex organic molecule—made possible by rapid improvements in underlying biological and information technologies including advanced data techniques.