Exploitation of microbiome science will unlock multiple commercial opportunities across a number of industries including life sciences. But there is a need for more effective data and process management to get the best from microbiome data, according to Dr Sven Sewitz, Director of Biodata Innovation at Eagle Genomics.
The microbiome comprises trillions of diverse bacteria and microbes that colonise our tissues, organs, and environment. It contributes to the formation of critical, specific ecological niches, where it carries out essential processes such as the breakdown of food and nutrients in the gut and can even influence the success of critical clinical treatment approaches such as immunotherapy. It is also being recognised that the microbiome can impact metabolic health and associated diseases like diabetes, which carry significant healthcare burdens. Our microbiomes have the potential to transform human health through nature-inspired innovation but also can impact the health and wellbeing of livestock and our environment, including areas like agriculture – with the potential to even help reverse climate change.
Advancements in areas such as human health can only be achieved with a more sophisticated understanding and application of the microbiome. To that end, global technology players have been addressing challenges related to microbiome data analysis with automation and artificial intelligence (AI)-based tools and platforms that can help perform contextualised data capture, curation, and reporting on microbiome data.
With potential to impact almost every aspect of our lives, the microbiome and its related innovations are an important component of a new wave of biology and/or nature-inspired innovation. The Bio Revolution could have a direct economic impact of up to $4 trillion a year over the next couple of decades, according to McKinsey research. In fact, bio-based innovations promise to provide regenerative capitalist 1 answers to the climate crisis, tapping into the application of nature’s laws and patterns of systemic health, self-organisation, self-renewal, and regenerative vitality to socioeconomic systems. The Bio Revolution promises natural solutions to challenges such as identifying and developing new therapeutics, treatment strategies and even how we approach patient stratification. Realising this potential is far from trivial though. Handling the complex multi-omic ‘big data’ relating to the microbiome to translate it to such innovations is a significant barrier to success. Researchers are often limited by time-consuming, inefficient data processing and management. A February/March 2020 Visioneers survey2 conducted with users of multi-omics data, including scientists, researchers and informaticians from the biopharmaceutical industry, revealed that researchers are struggling to manage such complex data. There is wasted time and effort associated with working with unreliable data, along with siloed and outdated processes which are difficult to transfer or scale-up to further stages of development.
However, solutions to these challenges are beginning to emerge. Innovative tools and platforms, combined with more standardised approaches to understanding and gaining insights from data are needed to unlock the promise of the microbiome to better support the accelerating Bio Revolution.
The gap between biological and data science expertise
Researchers can be challenged with incompatible data formats, measurement methods, and data transformation techniques. This can lead to missed opportunities for data integration at a higher level of abstraction. Currently, there is a lack of accessible informatics tools to enable such data integration. Again, the lack of standardised vocabularies, nomenclature and ontologies creates bottlenecks for higher level analysis.
Researchers may also find themselves unable to access the requisite tools and techniques for multi-omics analysis due to the gap between biological and data science expertise. For example, while machine learning is a powerful tool that can provide deep insights into microbiome research studies and identify patterns in datasets, it is a specialist branch of data science, and there is currently a dearth of data scientists available to carry out such work. This need is also likely to grow as more multi-modal and network-based analyses come online, including multiple disparate networks of data to integrate and interrogate.
Collaboration is key
A new generation of scientific data management and analysis platforms is needed for microbiome data and processes. There is opportunity in the application of microbiome science but the realities of multi-omics data management demonstrate the need for technology and platforms that can handle extreme volumes of information, process it and provide valuable outputs. These platforms must enable consumers of data – scientists, data scientists, product markets, researchers, members of legal and compliance and more – to understand and digest the output so complex data can be productively applied to real-time solutions. Bioinformaticians and data scientists must work together to build a new layer of data and information into their analysis, using advanced tools to handle the deluge of data. It is also imperative that robust systems are used to combine and manage datasets as a foundation for innovation and discovery for both data scientists and biologists.
AI plays a supporting role
AI can help reveal patterns in data that might not be detectable to human observers. In fact, organisations using active metadata, machine learning and data fabrics to dynamically connect and automate data management processes will reduce their time to data delivery, and impact on value by 30%, according to analyst Gartner.Other innovative approaches which are delivering great potential for understanding the environmental role of the microbiome include causal inference programming and network science. This is the process by which causative links are inferred from data. The causal inference programming approach is proving valuable to microbiome researchers, revealing associations between diverse data, allowing scientists to design new studies to delve deeper into root cause analysis.
A data-driven partnership
The microbiome industry is accelerating. Similar early-stage industries that have leveraged data effectively have gone on to deliver rewards that have benefited the economy and wider society. Examples include small molecule drug discovery as well as semiconductor process and product development. Each early-stage industry has benefited greatly from the adoption of standards, best practices, and sophisticated tools that were used to develop life-saving drugs or advanced electronics. The microbiome-based Bio Revolution offers hope for global challenges from soil degradation to human health, to food and farming. This could unlock financial gains to transform design and production capabilities across many sectors. To make this a reality, it is critical to enable a data-driven partnership between biology and technology, which could translate to a better quality of life for all.
Volume 23, Issue 2 – Spring 2022
About the author
Dr Sven Sewitz, Head of Biodata Innovation, is an experienced scientis with an interdisciplinary background. He gained his PhD in molecular and cellular biology from Oxford University and trained in translational biology as well as in bioinformatics and data science at the University of Cambridge. At Eagle Genomics, Sewitz heads the team in Biodata Innovation, focussing on metagenomic data analysis and graph learning technologies.