How can Big Data and AI contribute to more sustainability?

We asked the host of the track, Postdoc Douglas McCloskey from The Novo Nordisk Foundation Center for Biosustainability (DTU Biosustain), what the track on Big Data and AI at this year’s DTU High Tech Summit will focus on.

What can people expect to learn at your track?

“People can expect to learn about the current state of Big Data and AI in biotechnology, and how we and others in the community are seeking to incorporate Big Data and AI into the work that we do on a day to day basis.”

Why are Big Data and AI instrumental in the green transition? How do these technologies contribute to a more sustainable society?

“Bringing a single drug to market or engineering a single strain for the production of a single compound takes many years and lots of resources. For example, the cost to bring a single drug to market was estimated to be $2.6 billion as of 2019. What if instead of $2.6 billion for a single drug, it was $2.6 million for hundreds of drugs?

A major paradigm shift in the research and development would be needed for that type of change. In order to make that paradigm shift we need the ability to generate large volumes of -omics data, e.g. RNA sequencing, metabolomics, etc., rapidly and economically using fewer resources than it takes currently, and then utilise AI to turn that data into actionable insights quickly and reliable so that fewer validation experiments, that also consume resources, need to take place. 

Additionally, many products that do not yet exist could contribute directly to a more sustainable society. For example, what if there was a product that could use carbon dioxide directly from the atmosphere to make biodegradable plastics? Big Data and AI will most likely be critical to the development of the technology behind that product.”

How do we work with Big Data and AI here at DTU Biosustain?

“The majority of data that we work with here at the Center can best be characterised as small and expensive. This is in large part because the automation and informatics infrastructure for conducting large scale experiments does not yet exist and the per sample costs to generate large-omics data sets (e.g., RNA sequencing, metabolomics, etc.) is prohibitively expensive.

During the next few years, we will be working to build up our automation and informatics infrastructure and reduce the per sample costs through technology innovation and method development for omics data so that we can generate Big Data economically and sustainable. In the meantime, much of the AI work is dedicated to automating routine data processing tasks that are done manually, and prototyping new methods using publicly available data sets.”

How do you work with Big Data yourself?

I am currently working on developing the automation and analytics infrastructure needed to get Big Data economically. I am also working on several projects aimed at automating routine data processing tasks using deep learning and other machine learning methods and prototyping new deep learning and other machine learning methods that are meant to leverage large-omics data sets.”

What is your favourite Big Data/AI tool and why?

“I like graphics processing units (GPUs). They are the backbone of the current Deep Learning frameworks because of their ability to efficiently compute matrix multiplications. Without them, the current breakthroughs in AI that make the headlines would not be possible. More biotech- related; I also like mass spectrometers. There is currently not a better tool available for measuring small molecules, lipids, and peptides with greater resolution, specificity, throughput and speed.“

In which areas do you see Big Data and AI contribute to make consumers select greener choices in the future?

“The options for sustainable-conscious consumers are quite limited. Big Data and AI can help bring about the technology shifts that allow for providing greener choices. “