Astromech, the evolutionary biology AI company spun out of Colossal Biosciences, has closed an additional $20 million in funding, bringing its total raised to $60 million at a $3.8 billion valuation. The Dallas-based company plans to use the capital to expand its research team, scale its model training infrastructure, and launch forecasting pilots in health and biosecurity.
Bob Nelsen, co-founder and managing director of ARCH Venture Partners, led the round. PEAK6, NeoGenesis Capital, Builders VC, and CAZ Investments also participated.
Co-founded by Ben Lamm and Harvard geneticist George Church, Astromech builds predictive models that help researchers anticipate changes in living systems. The company compares its approach to weather forecasting: instead of predicting storms, it aims to help scientists get ahead of genetic bottlenecks, drug resistance, disease progression, and responses to changing environments.
"Biology runs the world and historically, we have only reacted to it," Lamm said in a statement. "We can describe biology in extraordinary detail, yet we still struggle to anticipate what comes next."
The science behind the model
Astromech draws on Colossal's genomic data resources, including a broad genome bank of extinct and living species. The goal is a platform that can predict where a genome, pathogen, or population may be headed, and what is driving those changes.
The company's first public demonstration mapped 46 longevity-associated genes across a time-calibrated tree of life - a species family tree scaled to when each lineage diverged. Researchers can trace how genes tied to longevity, cancer resistance, and cellular repair changed across millions of years, and compare those trends across entire groups of species.
Asian elephants, for example, evolved notable cancer-suppression mechanisms despite their size and long lifespans. Tasmanian devils, by contrast, remain vulnerable to a transmissible cancer that has devastated wild populations. Astromech's work seeks to explain why such divergent outcomes happen.
Church said the approach depends on tools that didn't exist a decade ago.
"Most of the variations that matter for complex traits … are regulatory rather than coding, so reconstructing the ancestral regulatory state, not just the ancestral protein, has the crucial explanatory power," he said in a statement. "That takes functional data across many species rather than sequence alone, and AI reconstruction cheap enough to run genome-wide."
Connection to Colossal
Astromech launched in 2025. Details of its work became public this spring. The company remains closely tied to Colossal, the de-extinction startup Lamm and Church founded in 2021.
Colossal made headlines for its mission to bring back the woolly mammoth and other extinct species. In January 2025, it became the first "decacorn" startup in Texas, with a valuation above $10 billion. In May 2026, the company said it hatched live chicks using a 3D-printed shell, a technique aimed at bringing back extinct birds too large for living surrogates.
What's next
Astromech described its current phase as deep research and development. Its platform and initial research pipelines are operational, but the company is still hiring across ancestral modeling, regulatory genomics, genomic inference, sequence reconstruction, metabolic modeling, and protein folding.
The next phase includes forecasting pilots with partners in health and biosecurity. Over time, Astromech plans to turn its predictions into early-warning systems and therapeutic research programs.
"Every genome carries a record of what changed, when it changed and the tradeoffs that followed - but almost none of that history is readable at scale today," Lamm said. "Astromech was built to that gap, and this funding allows us to further expand that work far more across the tree of life."
Why this matters for science and research professionals
Astromech's approach points to working with data that used to be the friction point in evolutionary biology. Researchers who want to test hypotheses about gene function evolution no longer need to assemble one species at a time; they can use existing functional genomic datasets across many species.
For scientists and researchers in academic labs or industry settings, the practical takeaway is in the attitude: AI for Science & Research is becoming a throughput tool for comparative biology. The stay relevant to this field will likely involve knowing how to work with large, multi-species institutional datasets and how to evaluate predictive models against real outcomes - not just sequence alignment.
The pilots in health and biosecurity will be worth watching. If Astromech's approach to forecasting change in living systems shows reliability beyond genomics, it could give researchers a new way to test ideas about drug resistance and disease progression before running slow, expensive experiments. For biology teams, patient follow: the models that predict where systems will break are the ones that survive in practice.
AI for Science & Research and broader Research resources may be useful for keeping up with how predictive models evolve into usable genomics tools.
Your membership also unlocks:
