Accelerating Drug Discovery and Development: Synergy Between Wet and Dry Lab
Why we invested in A-Alpha Bio
At Breakout Ventures, we invest at the convergence of biology, chemistry, and technology — what we call Creative Biosciences. Given the alignment of computational applications in drug discovery with our investment thesis, we’ve engaged with many remarkable founders and companies in this field. When evaluating these technologies, one of our first considerations is the differentiation of the platform. We want to understand the “secret sauce” — the unique and distinguishing factors that set a particular company apart. The source, generation, and input of data plays a critical role: Does the company rely on publicly available data? Do they pull their data from a private source, like a hospital system? Do they generate their own data through computational or wet-lab experiments? Or some combination of these approaches?
Often, a wet lab can provide that differentiated source of data generation. Throughout my time in research, I’ve held a bias toward experimental science over computational science. This bias stems partly from my training as an experimental organic chemist, but primarily from a sense of mystery surrounding computational sciences. For me, it resembled a black box of code, models, and simulations — not recognizing these inputs are often data from wet lab experiments. In retrospect, I realize that my binary mindset was shortsighted. At the end of the day, these are “tools in the toolbox” and should be treated as such and, most importantly, these tools can complement each other in profound ways.
With the advent of computational technologies, the realm of drug discovery has experienced a profound paradigm shift. However, it would be wrong to suggest that AI is single-handedly taking charge of drug discovery. Instead, AI has revolutionized our approach to drug discovery. Take, for instance, the case of Recursion: a company that acknowledges the value of their “dry lab”, but in the context of their wet lab and its immense data generation capabilities…
“The core principle of our approach to improving the scale and efficiency of drug discovery is to automate and integrate the wet lab to create massive empirical datasets of biology and the dry lab where we leverage machine learning, or ML, to unravel the complex patterns within our datasets. No static dataset could likely contribute meaningfully to solving disease biology; the secret is the iterative approach where the learnings of prior data inform the generation of new data, and the secret to that approach is to generate our own data in-house. Our dataset, fit for the purpose of machine learning, grows by approximately 80 terabytes each week, and thus algorithms can be improved exponentially faster than if applied to a static dataset.” — Recursion S-1 (March 2021)
For Recursion, the combined value of the platform transcends the sum of its parts. They even go so far as to assert that their “secret” lies in the iterative process that unfolds between the wet and dry aspects of their platform. Inspired by the power of iteration and the vast amount of data generated, we have been searching for the next groundbreaking company that embodies these principles, aiming to uncover transformative possibilities.
And we found them: meet A-Alpha Bio.
A-Alpha Bio is a platform therapeutics company, capitalizing on its unique ability to measure, predict, discover, and engineer protein-protein interactions (PPI). Breakout Ventures invested in A-Alpha Bio alongside Xontogeny, Madrona, Boom Capital, and several other existing investors.
We decided to fund and support A-Alpha Bio for three main reasons:
1) The technology. The core technology of A-Alpha Bio is made up of two distinct platforms: AlphaSeq (experimental) and AlphaBind (computational). AlphaSeq combines yeast biology and next-generation sequencing to quantitatively evaluate the affinities of millions of protein-protein interactions. The methodology begins by constructing two yeast surface display libraries: library a and library α. These libraries are subsequently mixed in a liquid culture, allowing for interactions between cells expressing complementary proteins. Upon contact, these cells bind and merge, resulting in the fusion of their genetic material. The frequency of cellular fusion directly corresponds to the affinity of the proteins expressed on the cell surface. AlphaBind serves as A-Alpha’s computational machine learning platform. At its core, AlphaBind utilizes data from AlphaSeq and leverages its algorithmic capabilities to enhance protein-protein interactions and gain deeper insights into their underlying mechanisms. By employing machine learning algorithms, AlphaBind can identify patterns, correlations, and trends within the data, enabling A-Alpha to gain a better understanding of the factors influencing the performance and outcomes of AlphaSeq. This not only enhances the understanding of protein-protein interactions but also provides insights into why certain interactions lead to the desired function while others do not.
For all the reasons mentioned above, we appreciate the platform embodies the iteration and large data set generation capabilities. A-Alpha Bio has been able to measure over 400 million PPIs to date and are on track to measure over a billion in the near future. Yes, you read that correctly — A BILLION. Unreal.
2) The market. Protein-protein interactions govern a massive part of understanding and treating disease, from new biologics to novel molecular glues. Combined, they represent markets that exceed over $300B in total market size. Large pharma companies and research institutions recognize that A-Alpha could revolutionize their discovery and optimization process and are lining up to work with them. A-Alpha has recently announced partnerships and collaborations with Bristol Myers Squibb, Lawrence Livermore National Laboratory, and Gilead Sciences, to name a few.
3) The team. In our diligence process, we were blown away by the team’s experience, performance, and determination to push A-Alpha Bio forward. The company was co-founded by David Younger and Randolph Lopez and spun out of the lab of PPI expert, Professor David Baker, at the University of Washington. In the end, it is the people that push through the hardest moments, build insanely dedicated teams, and are going to make the millions of decisions that determine if technology develops into a product that can change the world. We were captivated by their clarity of thought, aligned passions, and ability to mesh as a team. And the cherry on top: they are an absolute joy to work with.
We are thrilled to partner with A-Alpha Bio and work alongside them to deliver the future of therapeutics, powered by understanding protein-protein interactions at a massive scale.
To learn more about A-Alpha Bio and what they are building, visit aalphabio.com.
