Aller au contenu principal

Aqemia: in search of new therapeutic molecules

Innovation Article published on 17 April 2024 , Updated on 17 April 2024

Co-founded in 2019, Paris-based start-up Aqemia promises to revolutionise the search for new therapeutic molecules. Thanks to its technology uniting artificial intelligence and cutting-edge mathematics, the company can discover drug candidates 10,000 times more efficiently than its competitors.

Aqemia's story began at the École Normale Supérieure on rue d'Ulm (ENS), where its co-founder, Maximilien Levesque, was working as a researcher in statistical and quantum physics at the National Centre for Scientific Research (CNRS). "In 2017, I was working on an unsolved equation and found that it could be used for drug discovery. That's how I came up with the idea of creating Aqemia," he explains. The following year, he took part in and won the Ulm start-up competition, where he met Olivier Vaury, Finance Director at Amazon France. "He put me in touch with Emmanuelle Martiano, a CentraleSupélec graduate working as a strategy consultant for major groups. She became the co-founder of Aqemia and took on the position of COO," adds the CEO of the company, which was officially launched in 2019.

Technology combining artificial intelligence and mathematics

Aqemia's mission is focused on drug discovery through the highly efficient invention of therapeutic candidates in silico (trials developed using computerised calculations). Sixty people work on unique algorithms to design small molecules which will be used in the future to treat numerous diseases.

The start-up operates two types of technology: generative artificial intelligence (AI) and physical mathematics. "We use generative AI that invents molecules. It's not difficult to do ‭– there are already other companies doing it. The complexity lies in testing the activity of these molecules on their therapeutic target, and then testing their non-toxicity," explains Maximilien Levesque. This is where mathematics comes in. They run tests on an atomic scale to see whether the AI-generated molecules correspond to the therapeutic target, using binding free energy calculations. This virtual test passes on the results to the AI, which then learns and produces a new, more suitable substance. This process continues until the technology offers a molecule that is highly active on the therapeutic target, free to use (i.e. not patented) and non-toxic. Once these characteristics have been validated, the substance is manufactured and analysed by external service providers.‬

Unlike other technologies on the market, based on AI that need experimental data to train on, Aqemia tackles drug discovery projects from their earliest stage by generating its own data. This data is based on statistical and quantum physics algorithms patented by École Normale Supérieure-Université Paris Sciences & Lettres (ENS-PSL) and CNRS, for which Aqemia has an exclusive operating licence, worldwide and in all disciplines.

"Our virtual cycle takes a few weeks. The molecule is also synthesised in just a few weeks, and then tested," says the CEO. The start-up has launched more than a dozen in-house therapeutic projects, in oncology and immuno-oncology, but also in new therapeutic fields such as immunology, inflammation and central nervous system diseases.

To date, Aqemia has worked on three projects for which animal testing is underway. After the in vivo validation stage, the start-up will have to draw up the preclinical specifications for animals, and then submit a dossier to begin testing in humans.

Dynamic growth

Ever since it was launched, the start-up has benefited from invaluable support. The PSL-Valorisation department of the Université Paris Sciences et Lettres is in charge of technology transfer and intellectual and industrial property. As of September 2019, the Agoranov start-up incubator in Paris has also taken the start-up under its wing. That same year, Aqemia launched its first fund-raising round, raising €1 million from the investment fund Elaia - PSL Innovation Fund, which acquired a stake in the company. This was then followed by more fund-raising rounds every year: €10 million from investment group Eurazeo, €20 million from BPI France Large Venture, and finally €30 million raised at the end of 2023 from Wendel Growth, the venture capital subsidiary of French investment giant Wendel. A number of business angels, particularly from the worlds of public research, the pharmaceutical industry and tech start-ups, also contributed to this funding. With over €60 million raised, Aqemia expanded its team, developed its product and pipeline of in-house therapeutic programmes, and moved into a new office in Paris.

Always keeping a growth mindset, Aqemia then signed partnerships with major names in the pharmaceutical industry. In December 2021, the start-up announced its cooperation with French pharmaceutical group Servier for drug development. The contract was renewed in 2023 for the discovery of new molecules in immuno-oncology, on a reputedly difficult therapeutic target.

"In December 2023, we signed a commercial agreement worth €140 million with Sanofi, one of the ten biggest pharmaceutical groups in the world," says a delighted Maximilien Levesque. A previous collaboration with this group dated back to 2020, in the fight against SARS-CoV-2, the virus behind COVID-19, at a time when Sanofi's vaccine was still a long way off. Aqemia's CEO continues: "When we work with pharmaceutical companies, they present us with a therapeutic problem and we offer solutions using our technology."

In the coming months, the start-up hopes to scale up its in-house therapeutic projects and validate them in animal models. "Our goal is to start human clinical trials by 2025 or 2026," explains Maximilien Levesque. In addition to these objectives within France, Aqemia is aiming to gain a foothold in the US market in the near future. The company is looking to set up in Boston, Massachusetts, one of the world's most concentrated biotech and pharma hubs, attracting numerous investment funds and skilled engineers.

But to cover these ambitious projects, Aqemia needs more hands. "There's currently around 60 of us, but we're hoping to double our workforce by 2025. So, we're recruiting a lot of people with very scientific profiles and a background in biology and chemistry, even without any knowledge of IT. We need them to interpret our results. And of course, we're looking for people with backgrounds in software, machine learning, mathematics and atomic and computational physics," concludes Aqemia's CEO.