Bridging biosciences and AI; Artificial Intelligence in the Biosciences Network

11th August 2023

A new national network aims to spark science breakthroughs through bringing together bioscientists and artificial intelligence (AI) experts across the UK.

Quadram Institute Researcher Dr Dipali Singh is part of a group managing the new Artificial Intelligence in the Biosciences Network, funded by the Biological and Biological Sciences Research Council (BBSRC). We spoke to Dipali to find out more about the network, her career and the potential of AI in biosciences.

“Most biological problems are complex. In biosciences, we are producing more and more data than ever to try to understand these problems. But we need advanced tools to comprehend these large and complex biological datasets.

AI is a tool that can assist you and in the case of research that means assisting with data analysis and understanding the complexity of data.

As bioscience researchers we are already using AI tools that we don’t necessarily think of as AI. Principal Component Analysis (PCA) is widely used in biosciences and is an unsupervised machine learning approach, which comes under AI. We want to enhance the use of algorithms in bioscience even more, making biology a quantitative discipline.

Our new national network aims to bridge the gap between bioscience and AI, uniting two disciplines often seen as separate – biology and mathematics, on which AI is based.

Even as scientists, we speak different languages. An AI scientist or data scientist will speak in more mathematical terms while bioscience researchers will be talking in biological terms.

The network aims to bring both of these expertise together to foster interaction and exchange knowledge exchange, facilitating collaborative solutions for complex biological problems.

My research experience is a mix between biosciences and AI. My undergraduate degree was in Biotechnology at Kathmandu University in Nepal. Then in my masters I specialised in bioinformatics from the Indian Institute of Information Technology where I started to learn about machine learning and AI.

After my masters, I spent a year working for IBM where I learnt skills in databases which was really helpful when I decided to return to research for my PhD in Oxford Brookes University.

My PhD was in genome-scale metabolic modelling of diatoms. This was when I began incorporating mathematical models into a biological context.

I find the process of converting a biological problem into a mathematical equation fascinating.

Since my PhD, I have been dedicated to advancing metabolic models that merge mathematics and biology, with a specific focus on metabolism. Currently, at the Quadram Institute, my research involves studying metabolic readjustments in various conditions, both in individual organisms and within the microbiome, leveraging metabolic models.

How the network started

In the autumn of 2022, I saw that BBSRC were looking at how to support bioscientists in using AI.

I was pleased to see the research council putting the network together to bring together this expertise on a national level, from different backgrounds. They decided to start the national AI in Biosciences Network to do this.

I applied through a selection process to be part of the management team of the network and was successful. There are currently seven members in the management team, from different areas of bioscience along with AI experts . The network is lead by Professor Andrew French at the University of Nottingham.

The network encompasses both academia and industry, with support from institutes and industries such as Turing, Syngenta and NVIDIA. Involvement from industry is important because they can contribute their advanced expertise and computational power to tackle the complex real life bioscience challenges that we (bioscientist) bring to the table.

Activities to advance bioscience through AI

The approach we are taking in the network is to upskill bioscientists skills in AI. It doesn’t mean that bioscientists need to be experts in AI but at least they can understand how they can benefit from AI.

Conversely, AI experts may not know where to use their expertise or algorithms in a biological context. There are algorithms with potential for bioscience applications, but they might require refinement and adaptation to be effectively used in this field.

The network is community driven so workshops and events will be those that are identified by the community. Through the network we will also be funding pilot projects for early-career researchers to incorporate AI in biosciences. Any team of AI scientists and biosciences across the country will be able to apply.

The network also aims to host a data repository. For bioscientists, that might be uploading biological data while for AI experts that might be adding an AI model for others to use in the community.

The data repository will facilitate “sandpit” events, that we plan to organise, where bioscientists and AI experts can work together to solve a biological problem. We hope the network will be the go-to-place for people both in AI and biosciences and provide a platform that allows researchers to crowdsource solutions to biological data-related challenges and ask the community to solve the problem, similar to Kaggle.

One successful example of a crowdsource solution hosted by Kaggle is Netflix recommendation system. Now when you browse Netflix and receive recommendations, the AI algorithms developed through crowdsourcing are responsible for those suggestions.

Society’s say in how AI is used

AI is a hot topic people are talking about. I use aspects of AI in my research but I am also interested in the broader impact of AI on society and it is something the network is interested in too.

One of the objectives we have in the network is ethical and societal AI issues. The network is aware that there are societal and ethical concerns around AI. We hope to work with regulatory bodies, policymakers and research councils about the ethical and societal impact of AI.

We need policy to be up to date with AI technology and how it is used. International policy and regulation are very important for the future of AI and will decide where we end up.

The network is in its early days, starting in September. We will be sharing more about how you can get involved in the network over the coming months.

Here on the Norwich Research Park we have a hub of bioscience institutes and I am a point of contact for people in Norwich interested in knowing more about the Artificial Intelligence in the Biosciences Network.”

Related People

Related Research Groups

John Wain

Related Research Areas

A green background with an illustration of a gut full of microbes.

Food, Microbiome and Health

A black background with a spherical form of green and purple bacteria. Radiating out from the central spherical form and green and purple streaks.

Microbes and Food Safety