Synthesize Bio is an innovative biotechnology company that is making significant strides in the field of biological foundation models. Recently, Synthesize Bio announced a successful seed funding round, securing $10 million. This round was led by Madrona, with additional investments from companies like Sahsen Ventures, Inner Loop Capital, Point Field Partners, and AI2 Incubator.
The company’s models are used to simulate gene expression experiments, a crucial aspect of understanding genetic behavior and its impact on health. This financial backing will help accelerate the development of the company’s generative genomics models, making them more accessible to biopharmaceutical companies and researchers alike.
One of the significant achievements of Synthesize Bio is the development of their flagship product, the Generate Expression Model-1, or GEM-1. This is a foundation model designed explicitly for generative genomics. What sets GEM-1 apart is its training on one of the most thoroughly curated RNA-seq datasets ever compiled. A recent publication from Synthesize Bio highlights an exciting capability of GEM-1: it can generate in silico data that closely matches results from real laboratory experiments based solely on descriptions of experimental designs.
This marks a groundbreaking shift in the field and introduces the concept of “generative genomics.” In this approach, high-fidelity generative models not only replicate past laboratory results but also predict future experimental outcomes and clinical trial results, ultimately accelerating the pace of innovation in biotechnology and pharmaceuticals.
Dr. Bradley and Jeff Leek, PhD, are the founders of Synthesize Bio. Their vision is to empower researchers to accelerate drug discovery and other important applications that rely on gene expression data. By utilizing these advanced models, researchers can overcome the common challenges of limited and biased laboratory or clinical datasets, enabling them to make more informed decisions in their work.
Dr. Leek holds the position of Chief Data Officer at the Fred Hutchinson Cancer Center and is recognized for his pioneering work in RNA informatics. He has led substantial efforts to gather, normalize, and merge various RNA datasets from researchers worldwide, resulting in the largest integrated dataset currently available. His contributions to the field have not gone unnoticed, as he was recently named to the Time AI 100 list for 2025 for his initiatives in federating patient data among cancer centers.
On the other hand, Dr. Bradley serves as the McIlwain Family Endowed Chair and heads the Translational Data Science Integrated Research Center at the Fred Hutchinson Cancer Center. His research has revealed that dysregulation of RNA is a prevalent factor contributing to the initiation of cancer, and he has identified new therapeutic avenues for treating such cancers.
The initial performance of GEM-1 has been documented in a preprint available on bioRxiv. This report shows that GEM-1 successfully predicted the outcomes of laboratory experiments conducted after the model’s training data cutoff. Additionally, it demonstrated the ability to generate data from large clinical cohorts. This evidence supports the idea that studies conducted entirely in silico can yield results that are comparable to those obtained from traditional wet lab or clinical trials.
Looking ahead, the Synthesize Bio team is actively working on developing partnerships with biopharma teams to accelerate the process of drug development using their innovative foundation models.
Through these collaborations, they aim to mitigate risks throughout the clinical pipeline, from identifying reliable targets to simulating responses to therapies and optimizing the design of clinical trials. Currently, access to the GEM-1 foundation model is available through Synthesize Bio’s platform, along with support for R and Python API clients.
This accessibility will enable researchers and companies to leverage the power of generative genomics in their own work, fostering advancements in drug discovery and healthcare.
KEY QUOTES:
“Biopharma companies need rich, representative data to identify new drug targets, validate drug safety and efficacy, and power foundation models. Our models help scientists predict the results of new and currently impossible experiments to move their science forward much faster and much more cost effectively.”
Rob Bradley, PhD, Co-founder of Synthesize Bio
“Madrona’s investment builds on our longstanding thesis at the intersection of AI and life sciences. We believe Synthesize Bio represents a transformative shift in how biopharma and researchers efficiently discover and develop new therapies. Rob, Jeff, and the team are uniquely equipped to bring generative genomics into practice, and we are thrilled to partner with them.”
Matt McIlwain, Madrona Managing Director
