Two of the most consequential biotech outcomes of the past decade — Translate Bio's $3.2 billion acquisition by Sanofi and Recursion Pharmaceuticals' landmark IPO — share a common thread: early-stage investors who understood that the best science needs patient, conviction-driven capital long before the world catches up.
The journey from a university laboratory to a publicly traded company is one of the most improbable paths in all of business. In life sciences, the odds are particularly steep. Fewer than 10% of drug candidates that enter clinical trials ever reach patients. The capital requirements are enormous, the timelines stretch across decades, and the science itself is unforgiving — a single failed experiment can invalidate years of work. Yet within this landscape of long odds, certain companies break through in ways that reshape entire therapeutic categories. Translate Bio and Recursion Pharmaceuticals are two such companies. Their stories, examined side by side, reveal a set of principles about what it takes to build durable biotech enterprises from the earliest stages — and why the investors who back them at the beginning matter as much as the science itself.
Translate Bio began its life as RaNA Therapeutics, founded in 2011 out of research from Harvard, MIT, and the Scripps Research Institute. The founding scientific insight was deceptively simple: messenger RNA could be engineered to instruct human cells to produce therapeutic proteins, bypassing the need for traditional small-molecule or biologic drug manufacturing. At the time, mRNA was considered fragile, unreliable, and commercially irrelevant by most of the pharmaceutical establishment. The few researchers who believed in the platform — including pioneers at the University of Pennsylvania whose work on nucleoside modifications would later underpin the COVID-19 vaccines — operated in relative obscurity, publishing to small audiences and struggling to attract institutional funding.
The company pivoted in 2016, rebranding as Translate Bio and narrowing its focus to mRNA therapeutics for pulmonary diseases, particularly cystic fibrosis. This was a deliberate strategic choice: rather than chasing the broadest possible application of the mRNA platform, the team identified a disease area where inhaled mRNA delivery could offer a differentiated therapeutic approach. The early capital that supported this pivot — pre-seed and seed funding from investors willing to underwrite platform risk — proved decisive. Without it, the company would have remained a broad-spectrum research project rather than a focused therapeutic enterprise.
Recursion Pharmaceuticals, meanwhile, emerged from the University of Utah in 2013 with an equally contrarian thesis. Co-founder Chris Gibson, then a PhD student, had been studying the cellular effects of rare genetic diseases when he recognized that high-throughput imaging and machine learning could systematically map the relationships between genes, diseases, and drug compounds. The idea was to replace the traditional hypothesis-driven drug discovery process — where a scientist spends years studying a single target — with a data-driven approach that could screen millions of biological relationships simultaneously.
In 2013, this sounded like science fiction. Deep learning was still in its infancy. The computational infrastructure required to process millions of cellular images at scale did not yet exist in most academic settings. Gibson and his co-founders were told repeatedly that biology was too complex, too noisy, and too poorly understood for machine learning to contribute meaningfully to drug discovery. They raised their initial capital from a small group of investors who saw the convergence of cheap computing, advances in microscopy, and the explosion of biological datasets as a once-in-a-generation opportunity.
Both companies experienced moments where the trajectory of the entire enterprise hinged on a single decision or external event. For Translate Bio, the first critical inflection came in 2018 when Sanofi entered a collaboration agreement worth up to $805 million to develop mRNA vaccines for infectious diseases. This partnership validated the company's delivery technology in the eyes of the broader market and provided non-dilutive capital that extended the runway to pursue its pulmonary pipeline. The second, more dramatic inflection came in 2020 when the COVID-19 pandemic transformed mRNA from an academic curiosity into the most important therapeutic modality on the planet. Translate Bio's existing collaboration with Sanofi positioned it perfectly: the two companies rapidly advanced an mRNA COVID-19 vaccine candidate into clinical trials. By August 2021, Sanofi acquired Translate Bio outright for $3.2 billion, seeking to own the mRNA platform that would anchor its next decade of vaccine and therapeutic development.
For Recursion, the inflection points were more gradual but equally consequential. The first was the company's decision in 2017 to build its own wet lab infrastructure rather than outsource experiments to contract research organizations. This was expensive and operationally demanding, but it gave Recursion control over the quality and consistency of the biological data feeding its machine learning models — a competitive moat that proved nearly impossible for later entrants to replicate. The second inflection came when the company demonstrated, for the first time, that its platform could identify drug candidates for diseases that had no known chemical starting points. This was the proof of concept that separated Recursion from the dozens of other “AI for drug discovery” startups that had emerged by 2019 but lacked the integrated biology-and-computation infrastructure to generate genuinely novel insights.
The companies that endure are the ones that build platforms, not just products. A single drug candidate is a lottery ticket. A platform that generates candidates systematically is an engine.
Between 2010 and 2020, hundreds of biotech startups were founded on variations of the same two theses: mRNA therapeutics and AI-driven drug discovery. The vast majority failed or stagnated. What separated Translate Bio and Recursion was not simply better science — it was a combination of strategic discipline, team composition, and the quality of their earliest investors.
First, both companies had founding teams that combined deep scientific credibility with operational pragmatism. Translate Bio's leadership understood that mRNA delivery — getting the molecule into the right cells in the right tissues — was the binding constraint, not the mRNA construct itself. They organized the entire company around solving the delivery problem, which meant hiring lipid nanoparticle chemists and formulation scientists rather than expanding the molecular biology team. Recursion's founders understood that AI in biology was only as good as the biological data it learned from, so they invested heavily in generating proprietary datasets at a scale no academic lab could match.
Second, both companies adopted what might be called a “platform-first, product-second” strategy. Many of their competitors rushed to advance individual drug candidates into the clinic, hoping that early clinical data would attract the partnerships and capital needed to survive. Translate Bio and Recursion took the opposite approach: they invested years building robust, repeatable technology platforms before selecting their lead clinical programs. This required more patience from investors and longer periods without the clinical milestones that typically drive biotech valuations. But the result was that when they did advance candidates, the pipeline behind those candidates was deep and the platform itself had independent strategic value to potential acquirers and partners.
The pattern we observe in both companies reinforces a thesis that Legacy Venture Capital has held since its founding: the most consequential decisions in a biotech company's life are made in the first eighteen months, when the science is unproven, the team is small, and the capital is scarce. GP Diane Shao's own experience underscores this. Her foundational research on Pentraxin — a family of proteins involved in tissue remodeling and fibrosis — led to the creation of Promedior, which was ultimately acquired by Roche for $1.4 billion. That outcome was not the result of late-stage capital optimization. It was the result of an early scientific insight, pursued with discipline, supported by investors who understood the biology well enough to tolerate the uncertainty.
Pre-seed and seed capital in life sciences serves a fundamentally different function than it does in software. In software, early capital buys time to find product-market fit. In biotech, early capital buys the experiments that determine whether a scientific hypothesis is commercially viable at all. The quality of those experiments — the rigor of the assay design, the choice of disease model, the decision about which data to generate first — determines everything that follows. Investors who lack scientific fluency at this stage often push companies toward premature clinical development or unfocused platform expansion, either of which can be fatal.
In life sciences, the first check is not just capital — it is a signal to the entire ecosystem that someone with scientific judgment believes this work deserves to exist as a company.
Taken together, the Translate Bio and Recursion stories offer a set of durable lessons for scientists considering the transition from academia to company-building. These are not abstract principles — they are patterns that recur across nearly every successful biotech outcome we have studied.
The therapeutic and technological landscape has shifted dramatically since Translate Bio and Recursion were founded. mRNA is no longer contrarian — it is consensus. AI-driven drug discovery has attracted tens of billions in venture capital. The question for early-stage investors today is not whether these approaches work, but where the next wave of differentiated platforms will emerge.
At Legacy Venture Capital, we see the same pattern of opportunity forming in several adjacent spaces. Programmable biology — including base editing, epigenetic therapies, and engineered cell therapies — is at the stage where mRNA was in 2012: scientifically validated in academic settings but not yet translated into scalable commercial platforms. Computational biology is evolving beyond image-based drug screening toward foundation models trained on multi-omic datasets that can predict drug behavior across entire biological systems. And the convergence of synthetic biology with manufacturing — the ability to produce complex biological molecules at industrial scale — is creating opportunities that look structurally similar to the early days of mRNA delivery optimization.
In each of these areas, the founders we seek to back share characteristics with the early teams at Translate Bio and Recursion: deep scientific expertise, a clear understanding of the translational bottleneck in their field, and the discipline to build platforms rather than chase the nearest clinical milestone. They are often working in areas that the broader venture market has not yet recognized as commercially viable, which is precisely when early-stage capital has the greatest impact and generates the greatest returns.
Diane Shao's journey from Pentraxin research to the founding of Promedior — and its eventual acquisition by Roche — illustrates the fundamental conviction that animates Legacy's investment approach. The most transformative biotech companies do not begin with a business plan. They begin with a scientist who sees something true about biology that the rest of the world has not yet understood, and an investor willing to bet on that insight before there is a shred of clinical evidence to support it.
The lab bench is where billion-dollar companies are born. Our job is to be in the room when the founder first realizes that their science can change medicine — and to give them the capital and conviction to prove it.
The stories of Translate Bio and Recursion are not anomalies. They are templates. The next $3 billion acquisition and the next landmark biotech IPO are being seeded right now, in laboratories where brilliant scientists are wrestling with problems the market does not yet understand. The investors who find them first — who have the scientific fluency to recognize the opportunity and the patience to support it through years of platform development — will capture returns that are impossible to generate any other way. That is the thesis. And the evidence, from mRNA to AI-driven discovery to pentraxin biology, continues to bear it out.
See the full Legacy Venture Capital portfolio and track record.
View Investments