Bits In Bio Interview: Elise De Reus PhD, Co-Founder Cradle.bio
AI in Bio Updates - From and Around the Bits in Bio Community
We are thrilled to kick off the first entry in our new interview series with the brilliant Elise De Reus PhD cofounder of Cradle.Bio
Q: Can you give us a quick overview of Cradle and what problem you're solving?
A: At Cradle, we're addressing a fundamental challenge in protein design. Traditionally, this field has relied heavily on trial and error, which, given the vast design space, leads to long development timelines and high failure rates. We're building a software platform that puts generative AI at the fingertips of pharma an non-pharma R&D teams, helping them hit their development goals faster.
Our platform is designed to work for multiple stakeholders, from associates to scientists, and it gets better with use. We've partnered with a wide range of organizations, from industry giants like J&J and Novo to earlier-stage startups. Importantly, we use a flat-fee billing model rather than charging royalties, as we aim to be part of the broad tech stack and empower as many people as possible across the industry.
Q: That's fascinating. Can you tell us a bit about the founding journey of Cradle?
A: Certainly. We started Cradle in October 2021. My co-founder Harman and I came from high throughput experimentation at Zymergen and DSM, while our other co-founders, Stef and Eli, brought their machine learning expertise from Google. Our fifth co-founder, Jelle, is exceptional at building usable software.
What was really helpful in our discovery phase was falling in love with the problem, not the solution. We looked at the challenges in the biotech field and the capabilities of machine learning, seeking to explore the intersection of these fields through an awesome and highly-usable digital product.
Q: Where does Cradle fit within the life sciences ecosystem?
A: If you think of the drug development pipeline, we're not involved in the initial target identification phase. Where we really shine is in the discovery and optimization phase, particularly in hit-to-lead identification and lead optimization. This is where drug companies typically spend the most money, and where time is critical. We help our customers develop better candidates in less time to increase chances of success in the preclinical stage, where there are typically only a small number of candidates in the running, and beyond.
In the synthetic biology space, we're working with companies who use the Cradle Platform for a range of different applications. Often to engineer better enzymes, moving beyond starting points discovered in nature.
Customers who use the Cradle platform test the ML-generated protein designs in in-house wet lab experiments, or at a CRO or cloud lab of their choice. Cradle can be seen as the ML-guided protein design tool in the biotech stack, which would be used alongside ELN and LIMS software, DNA synthesis vendors, and other service providers.
Q: It sounds like you've done a great job blending biology and AI. How has this interdisciplinary approach benefited Cradle?
A: It's been incredibly valuable. Having expertise in both areas helps the two sides of the biotech coin understand each other better. An important thing to note is that we invested in an in-house wet lab right from Cradle’s start, which allows us to test protein designs from our ML models at a very quick turnaround time.
This helps the Cradle team appreciate the value and complexity of generating high quality wet lab data. Our biology team runs a wide variety of test projects, paying close attention to controls and test cases that allow us to build robust models. Conversely, our ML team gains insights into the nuances of biological data generation. It allows them to build the Cradle Platform in such a way that it can deal with things like batch-to-batch variation, which might be caused by differences between operators or switching out a buffer.
An interesting feature of our system is that it allows direct expert input into the models, essentially encoding human intelligence. This helps users derive new scientific hypotheses through iteration with the algorithms. It's really a conversation between human expertise and machine learning.
Q: Where do you see the field of TechBio evolving?
A: I believe we're in an early adopter phase when it comes to utilizing AI in biotech R&D. There's a lot of FOMO - many companies are trying to figure out what they should or could do with AI in biotech. This often involves make-or-buy decisions; do you invest in hiring an ML engineer, explore commercial offerings, or both? At Cradle we're doing a lot of education around the potential applications as well as the ways to think about this decision.
I expect protein AI models - both predictive and generative - to become more mainstream in the coming years. The field will need to continue testing different models in different applications, and observing what drives value in individual workflows.
For example, in some R&D workflows it is still common to start with screening a large library (eg. an immunization campaign, or a site saturation mutagenesis library). But we’re moving towards a world where generative models can supplement or even replace that phase.
I'm also excited about the potential for easier 'prototyping' of early chemical product ideas. Imagine designing a protein, outsourcing data generation, and then doing a super lean small-scale test reactor run. Cloud labs can play a big role here too.
Q: What are you most excited about in your work?
A: What I love most is the visibility I get into the huge array of problems people are working on. It's truly inspiring to see what can be done with biology - from new approaches to curing cancer to reducing climate impact and ensuring equitable drug access.
I'm excited about driving down the cost of early research phases and making it easier to ideate and prototype bio programs. We're seeing benchtop DNA synthesizers becoming rapidly faster and cheaper. Combining these with better design tools and multi-modal models will accelerate the entire industry.
I'm particularly excited about advancements in vaccines and biopreparedness. The speed of response and potential for equitable access in this area are really compelling.
Q: Finally, what advice would you give to those looking to enter the TechBio field?
A: I'd say commit to developing expertise in a specific area, whether it's biotech, AI, or another relevant field. Pursue deep knowledge without necessarily knowing where it will lead. Then, look up and out to see where your contribution can make the most impact.
At Cradle, we value what we call the "Humble Steward, Generous Teacher, Curious Student" approach. It's about being open to learning from others while also sharing your own knowledge.
Lastly, I'd encourage people to engage with thought leaders in the industry. There are some great blogs and newsletters out there that provide insights beyond what you'll find in academic journals. Authors like Nico McCarthy and Elliot Hershberg offer valuable perspectives on the industry.
The field of TechBio is evolving rapidly, and there are incredible opportunities for those who are passionate about making a difference at the intersection of biology and technology.
Reach out to Elise De Reus PhD on linkedin, and reach out to Cradle.Bio to learn more about their evolving services and offerings