Eden Bio: Using machine-learning to improve protein production.

Written by Aonghus Topham, Edited by Caroline Babisz, Charlotte Pugsley and Charles Leach.

Proteins are essential across many aspects of our lives, from food and therapeutics, to the industrial enzymes currently in use to degrade plastics and other undesirable climate-damaging products out there. Eden Bio was founded in 2022 with the vision of using machine learning to optimise how these proteins get produced. Using their suite of computational and wet-lab tools, the team quickly screen the many possible genetic modifications that can be made to fungal strains to improve how much protein the strain can produce, helping their customers improve their yield and get their exciting new protein products to market quicker.

There are many different applications of the  proteins Eden Bio optimise the production of, but one of the most exciting is Precision Fermentation. In this process novel food products are manufactured through the brewing of genetically modified microorganisms, such as yeast - much in the same way beer is brewed today! The goal is to create a “Closed loop process”, where novel proteins can be made at a reduced cost. Many of the companies attempting this method are struggling to get their protein yield high enough to make it commercially viable - enter Eden Bio and their machine learning tools.

“Biology is so complex, that to understand all the information present, we need machine learning to cut through to the important bits, helping us make the connections we humans can’t on our own,” Dr Christopher Reynolds

Eden Bio’s origins

Eden Bio’s founder, Dr Chris Reynolds, has a rare background intersecting the disciplines of synthetic biology and machine learning. In his early career, Chris worked under two of the UK’s founding fathers of synthetic biology, Professors Paul Freemont and Richard Kitney, from whom he gained the insight he would later need to co-found his first company, Better Dairy. During his time as CTO of Better Dairy, Chris was able to demonstrate the potential of using machine learning to improve protein yield, which became the foundation of his new company, Eden Bio, when he exited Better Dairy after helping the company to its £16M Series A.

Chris’s goal with Eden Bio was to maximise the potential of combining the worlds of synthetic biology and machine learning, creating an impact across protein production. With the help of Eden Bio’s COO, Dr Rachel Shaw, Chris set up Eden Bio, and rapidly progressed the company through the choppy waters early stage start-ups must face, from building the team and raising their Seed round of £1m, and bringing in their first customer.

2022 was a difficult year for start-ups, with the recession, high inflation and market corrections seeing biotech investors less inclined to take the risks associated with early-stage companies. But Eden Bio is not your average start-up. With the knowledge and technology Chris was able to bring across from Better Dairy, coupled with Rachel’s fundraising experience, Eden Bio have been able to move at a much faster pace than start-ups are usually able to within their first year.

“There are lots of start-ups in this space with game-changing technology. However, they have many hurdles to overcome. We’re fortunate that we had Chris’s expertise to help us through those challenges,” Dr Rachel Shaw

How Eden Bio uses machine learning to guide strain engineering

To engineer a microorganism to produce a new protein, the gene encoding for the desired protein must be inserted into the genome of a host microbe. Following this, the host cell’s genome is further edited to increase the protein’s yield,  in ways that make the microbe more robust and able to survive the manufacturing conditions found inside an industrial fermenter.

There are infinite possible combinations of genetic tweaks that can be made to these microbes, and Eden Bio’s unique machine-learning platform provides the key to simplifying this laborious process. By testing the machine learning’s suggestions in high-throughput, they can quickly iterate through the search space, identifying the edits resulting in the highest possible yield.

With the help of Eden Bio, companies in this space can use machine learning to get their strains to the yields they need for commercial production sooner than would otherwise be possible.

Eden Bio’s proprietary machine learning platform is not limited to the sustainable food market either. Chris describes Eden Bio’s technology as “protein agnostic” with the potential to penetrate new markets. This means that they can use their technology to help any company looking at increasing a protein’s yield, and the team encourages any companies or academics interested in utilising this exciting new technology to get in touch.

We sat down with the CEO of Eden Bio, Dr. Christopher Reynolds and COO, Dr. Rachel Shaw to explore more about Eden Bio and what the future may hold.

The growing problem of sustainable food production

With the human population expanding at an exponential rate, it has become more apparent in the last decade that we are in dire need of new, sustainable food solutions. The holy grail is to create commercially viable, ethical, sustainable, cheap, tasty and nutritious food products.

With the nearly unlimited potential of precision fermentation, this industry has seen massive growth in the last 5 years, with more big-name investors turning their attention to this space. As a result, a number of exciting biotech companies are investing in this technology.

Eden Bio is perfectly situated to take advantage of this boom through combining biological techniques with their machine learning platform, and identifying the best genetic solutions for strain optimisation. This will allow this specialist company to fill the gap within the precision fermentation space of guaranteed strain optimisation.  

What are the main challenges in coupling Machine Learning and Synthetic Biology?

The massive variation within and between proteins makes the coupling of machine learning and synthetic biology akin to making the unpredictable, predictable. There are as many peptide variations within a sixty-amino-acid long stretch of protein as there are atoms in the universe.

One huge challenge is obtaining the amount of data needed for a machine-learning platform used to guide prediction points. With so many different outcomes, the sheer volume of initial data-points needed is almost unfathomable. Eden Bio’s solution to this problem is to use the power of deep learning to make connections in multiple dimensions.

Only now is science able to generate the volume of data needed in a commercially viable way, mainly due to the cost of sequencing and synthesising decreasing by several orders of magnitude in recent years. With the advances in computing power and deep learning in the last 5 years, machine-learning models able to reliably predict the complexity of biology are now possible.

What does the future hold for Eden Bio?

In November 2022, Eden Bio announced the completion of a £1M Seed funding round, paving the way for the company to expand their R&D capacity. This was led by SynBioVen, a new synthetic biology fund whose board consists of Professors Paul Freemont, Richard Kitney, Lord David Willetts (Chair of the UK Space Agency, former Minister for Universities and Science, and creator of the Synthetic Biology Leadership Council), and Sir David Harding (Founder and CEO of Winton Ventures).

It's clear that Precision Fermentation is poised to shake up the protein production industry, and with their next generation platform,Eden Bio is leading the charge. We're excited to follow their journey and support them along the way!

Recently, Eden Bio completed our network accelerator programme, Innovators Club. Run from September through to December each year, Innovators Club offers a series of workshops, mentoring and events where early stage start-ups working in similar areas meet each other and share common hardships. Dr. Reynolds described the experience as “A great insight into the world of investing, and a useful connector to other biotech companies in the space”.

‍