• Article
  • Science

LabGenius: AI-driven synthetic biology

London-based SynBio startup LabGenius integrates gene synthesis, machine learning and robotic automation to engineer better and entirely new proteins.

June 11, 2018

Bridging scientific computing, engineering biology and next generation genome sequencing is at the heart of Earlham Institute’s mission toward decoding living systems. At a recent workshop, we heard from Harry Rickerby of LabGenius, a startup driving automated, AI-driven biomolecule evolution.

Synthetic biology applies the concepts of engineering to building novel biological systems from their base components, a process that can be automated in labs such as EI’s DNA Foundry, which is home to a suite of state-of-the-art machinery allowing large-scale assembly of biomolecules.

London-based LabGenius, who have been awarded money from the BBSRC, are integrating gene synthesis, machine learning and robotic automation to engineer better & entirely new proteins with a range of applications from advanced materials to personal care. We spoke to Harry Rickerby, Chief Scientific Officer, about LabGenius and his recent visit to EI.

What got you keen on SynBio?

I first learned about Synbio during my undergraduate in Biology at Imperial College. I was fascinated by the idea of applying my knowledge of biology to applications in the real world. Synbio offered the tantalising opportunity to make things with biology, to program complex genetic circuits, to engineer new life from scratch. We’re not even close to reaching the boundaries of what we can do in this space, so the opportunities are massive.

One important area where LabGenius can apply their system is in the production of protease free pharmaceuticals.

What is different about your approach to SynBio?

LabGenius’ primary goal is to engineer new protein components with novel, enabling functionality. What sets us apart from others in the protein engineering space is our platform: EVA. We’re developing an entirely AI-driven ‘closed loop’ platform. This allows us to greatly accelerate the search of huge areas of sequence space without introducing human error and bias.

So how does it work, AI-driven protein engineering?

The process entails 4 main steps: First, large DNA libraries are designed to explore an interesting area of sequence space. These DNA libraries are then constructed in the lab using our own proprietary DNA library assembly method. Next, the DNA libraries are screened using ultra-high throughput screening methods to identify proteins with extraordinary properties. Finally, this data is fed into machine learning algorithms to unpick the genetic design rules that underpin strong candidates. This narrows the search space for EVA’s next cycle. We keep iterating around until we find a protein with the properties that we need.

What are the most exciting products you can make with your pipeline?

Currently we’re really excited about our work in the Pharmaceuticals space, specifically generating protease stable biological molecules. These could have a huge impact in allowing us to deliver biological drugs orally.

What did you think of the snazzy kit at EI?

The kit at EI is brilliant! I was particularly impressed by the banks of next generation sequencers there, and I’d certainly love to have a play around with the acoustic liquid handlers!

What sort of things are you looking to collaborate on in the future?

With academia, we’re really interested in any new high-throughput screening techniques that are being developed - we’re particularly interested in applications of microfluidics in protein screening. We’re also big advocates of cell-free biology [Susan Duncan and Quentin Dudley can help there], and we’re interested in setting up collaborations with groups who are doing pioneering work in this area.

Are you looking to incorporate SynBio into your home brewing any time soon?

That’s confidential… but seriously, that’s confidential.

Open quote marks

One of my favourite natural proteins is resilin. It’s an elastic protein used by insects to jump to to 38 times their length - it stores up to 97% of its energy, with only 3% lost as heat. I’d love to be able to produce this protein at scale… imagine the bouncy balls!

Closing quote marks

What is your dream synthetic organism, or product?

One of my favourite natural proteins is resilin. It’s an elastic protein used by insects to jump to to 38 times their length - it stores up to 97% of its energy, with only 3% lost as heat. I’d love to be able to produce this protein at scale. It would have some really cool applications in biomedicine, specifically in long term implants, but also… imagine the bouncy balls!

Do you have any advice for someone starting up in the startup world?

My advice would be to stop waiting. Too many entrepreneurs coming from a scientific background are held back by their desire to perfect their invention before starting a company. Once you’re sure about your idea, get out and raise some money to pursue it. If you don’t, you’ll be left in the dust by our slightly more bold American cousins! It sounds obvious, but your progress will massively accelerate once you have more resources.

Synthetic biology at the Earlham Institute

With our recently launched DNA Foundry, spearheaded by Dr. Nicola Patron, EI are helping to rebuild life from its very building blocks. With an expanding team, our synthetic biology capabilities will help us to tackle a swathe of important biological issues, from health and food security to complex manufacturing. The DNA foundry is a high-throughput, automated laboratory that supports the activities of researchers on Norwich Research Park and beyond to perform large scale construction of synthetic biological parts.

Article author

Peter Bickerton

Scientific Communications & Outreach Manager