Thomas Ybert and the cost of waiting

Thomas Ybert could design an experiment faster than he could obtain the material to do it. He was working on yeast engineered to produce jet fuel, a task with a pleasingly extravagant destination and an unglamorous obstacle. The DNA needed to change the yeast had to be made. Until it arrived, the next idea remained an idea. In the Bay Area, surrounded by businesses that could revise software almost at will, he was learning the peculiar price of waiting.

The frustration became DNA Script, founded in 2014 with Sylvain Gariel and Xavier Godron. Its eventual product, SYNTAX, brought enzymatic DNA synthesis onto the laboratory bench. Ybert spent a decade as the company’s founding chief executive. In October 2024 he became Chief Scientific Officer, making scientific development his principal job again. The thread connecting those decisions is a researcher’s wish to shorten the distance between thinking of something and being able to try it.

The jet-fuel problem had four letters

Ybert’s training took him through École Polytechnique and a biotechnology doctorate. Between 2007 and 2010, his doctoral work placed him in Sanofi’s research and development operation, working with yeast. He then joined Total’s New Energies division and helped establish its bio-jet-fuel collaboration with Amyris. That took him to the San Francisco area for two years, before further work in France. His route into entrepreneurship ran through industrial research, where an elegant idea still had to survive an actual production process.

At Total and Amyris, the goal was to make yeast produce a useful fuel. Doing that involved changing genetic instructions and testing the consequences. Ybert described a mismatch: design could move quickly, testing could move quickly, yet building the necessary DNA took weeks or months. An experiment has a timetable of its own. An order book has another. The researcher gets to discover which one wins.

“So, I dreamed about a genome printer,” he said. The appeal of the phrase is its domestic simplicity. A printer is something you expect to obey a file, produce a physical object and let you get on with your day. For a scientist working with synthetic DNA, that ordinary expectation was ambitious. He wanted a more direct relationship between a sequence on a screen and a molecule in a tube.

“So, I dreamed about a genome printer.”

Thomas Ybert · 2021

A company before a comfortable answer

Ybert met his future co-founders through Total. They began giving the project personal time in 2013. During 2014 he left the company to devote himself to DNA Script. Early support included funding of €150,000, a collaboration with Institut Pasteur and patent filings. The business had a direction before it had the fully functioning machine that would justify the direction. Anyone who has ever bought office equipment might reasonably prefer the reverse order.

The founders’ work also involved the Institut Pierre-Gilles de Gennes. These relationships mattered because the proposal crossed several kinds of expertise. There was an enzyme to engineer, a process to control and, eventually, an instrument to make that process usable. Ybert has consistently described the contribution of research partners as part of the company’s development. Inventing a tool for other scientists required spending time with other scientists.

In 2016, the Fondation Norbert Ségard recognized him with its young engineer entrepreneur prize. His explanation for entering was revealing: he wanted independent scrutiny as well as resources and credibility. He also stressed the importance of admitting one’s limits and finding people who could help overcome them. For an enterprise built around a new way to manufacture DNA, that is a rather useful starting attitude. Molecules are rarely impressed by certainty.

The Christmas present was ten bases long

For a period in the company’s early development, the team struggled to extend its DNA sequences beyond a few nucleotides. Looking back through his laboratory notebook, Ybert noticed an enzyme variant worth revisiting. He asked colleagues to test it before flying to California for Christmas with his in-laws. When he landed at San Francisco airport on December 24, a picture of the result was waiting: a clean ten-nucleotide sequence.

“This is the best Christmas present,” he recalled. The picture helped advance conversations with Illumina Ventures about a Series A investment. It is a compact account of what early scientific entrepreneurship can feel like: a large ambition briefly depends on a very small result. Ten bases were enough to make the next conversation possible. The notebook had earned its luggage allowance.

That scale deserves attention. DNA’s alphabet has four letters, A, T, G and C. A nucleotide is one of the units carrying those letters; an oligonucleotide, or oligo, is a short chain of them. Making a chosen chain requires getting the order right repeatedly. The difference between three units and ten can therefore be evidence that a process is becoming controllable. Size alone does not tell you what a result is worth.

Teaching an enzyme some manners

The enzyme behind this approach, terminal deoxynucleotidyl transferase, can add nucleotides without copying an existing DNA template. That makes it interesting for writing a sequence from scratch. It also presents a problem: an enzyme willing to keep adding bases needs a way to stop at the appropriate moment. Otherwise, the sequence you requested and the sequence you receive can part company rather quickly.

One method uses a temporary blocking group on a nucleotide. After that unit joins the growing chain, the block prevents another addition. Removing the block permits the next cycle. The principle sounds tidy in a diagram. Building reliable enzymes and repeatable conditions around it is the work. A machine must perform the sequence of operations consistently, across many different requested strings, without asking the user to become its personal biochemistry department.

By 2019, DNA Script reported enzymatic synthesis of a 200-nucleotide strand. That laboratory milestone helped establish a route toward products. It also illustrates why development numbers need labels: the length demonstrated in an experiment and the specifications of a commercial instrument answer different questions. A scientist buying a printer needs to know what it will deliver in ordinary use, including preparation, processing and the condition of the finished material.

A printer has to finish the job

The commercial launch came on June 15, 2021. The first SYNTAX system could synthesize 96 oligos in parallel, each up to 60 nucleotides long. The announced timings were about six hours for 20-nucleotide oligos and about thirteen hours for 60-nucleotide ones, with less than fifteen minutes of setup. Overnight production offered a particularly understandable bargain: leave instructions before going home, return to material for the next experiment.

The system combined the printer, reagent and consumable kits, and software. It also purified, quantified and normalized its output. Those quieter verbs explain much of the product. A strand that has been synthesized still needs to arrive in a usable condition. The scientist wants to start the experiment, rather than spend the morning completing the printer’s homework. Automation earns its place by absorbing those jobs.

Local production also changes the handling of a laboratory’s sequence information. At launch, DNA Script emphasized the ability to retain control over proprietary designs. Speed was part of the proposition, alongside control over scheduling and access to materials. Ybert’s early irritation had developed into a set of practical purchasing reasons. A useful invention has to make sense to someone responsible for tomorrow’s work and this year’s budget.

The Atlantic was part of the plan

In 2019, DNA Script announced a US office in South San Francisco to support commercialization. The location returned Ybert’s project to the region where his original frustration had taken shape. It also put the business near instrument makers, researchers and potential customers. A tool can be conceived in France and still require people on another continent to sell it, support it and explain why a laboratory should change its routine.

That geographical spread continued. In March 2026, the company announced distribution agreements covering Latin America, Japan and South Korea. In July, an agreement with Imperial Life Sciences extended local distribution to India. The underlying proposition remained recognizable: researchers far from centralized DNA manufacturing should be able to produce material within their own laboratories. Distance is an expensive participant in an experiment, especially when every revision starts another delivery.

The India announcement described a platform capable of producing up to 384 oligonucleotides in parallel. Five years earlier, the launch figure had been 96. These are dated product descriptions, rather than promises about every configuration. They show a business continuing to work on the practical reach of the printer Ybert once imagined. His personal story now sits inside a much larger network of operators, distributors and users.

Another job, the same impatience

When Marc Montserrat became CEO in October 2024, DNA Script reported more than €265 million in equity financing, over 150 patent applications and more than €20 million of income from SYNTAX and collaborations since inception. Those figures describe accumulated company activity. They also give the handover some scale: Ybert was changing roles within an organization that had moved a considerable distance from its early experiments.

He wanted to “focus more on the scientific innovation that defines our company.” The wording is useful because it names what he intended to do next. His move to Chief Scientific Officer put research and innovation at the center of his responsibilities, while Montserrat took on commercial leadership. A founder’s attachment to a problem can survive a change in the job title attached to his name.

The original wish remains easy to grasp. Ybert wanted more chances to test an idea before momentum drained away. DNA Script turned that wish into enzymes, instruments and a business. His story gives waiting a physical shape: an unmade sequence between a design and an experiment. The satisfaction comes when that sequence is finally in the tube, and the researcher can ask the next question.