Exploring Untapped Natural Chemistry for Future Medicines

In an exclusive interview with BioPharm International®, Guillaume Vandenesch, CEO and co-founder of Generare, a Paris-based techbio company, discusses emerging challenges in artificial intelligence (AI)-driven drug discovery. In particular, he emphasizes the need for more diverse and biologically relevant molecular data.

According to Vandenesch, current AI models are largely trained on narrow molecular datasets, limiting their ability to generate truly innovative chemistry. While models can produce novel compounds, he notes that novelty does not always equate to meaningful biological innovation or therapeutic relevance.

The company’s approach, as Vandenesch explains it, centers on leveraging natural compounds produced by microbes, which have evolved over billions of years under selective pressure to interact with biological targets.¹ These molecules offer a largely untapped source of chemically diverse and biologically validated data for training AI systems, he says.

How can natural chemical space improve AI drug discovery models?

The company aims to expand the data available to AI models by exploring microbial chemical space, of which only a small fraction has been studied. Historically, naturally derived compounds have led to the development of more than 500 drugs, including antibiotics and cardiovascular therapies.² However, Vandenesch emphasizes that this number represents only a small portion of the total available chemical diversity.

“And if 500 drugs gave us, I mean, were written, from 3% of the chemical space, what can we do with the remaining 97%?” he asks.

In early April 2026, Generare raised €20 million (US$23 million) in a Series A financing round, which is intended to scale its evolution-based molecular discovery platform. This investment speaks to growing investor interest in data-driven approaches for expanding chemical space for drug development.³

Vandenesch’s perspective poses a broader industry challenge of improving AI performance by enhancing the quality and diversity of input data. He notes that building comprehensive datasets rooted in natural chemistry may be a foundational step toward enabling AI systems to generate clinically relevant drug candidates and advance future therapeutic innovation.

About the speaker

Guillaume Vandenesch, CEO and Co-Founder, Generare

Vandenesch leads Generare, a Paris-based techbio company generating novel, high-quality molecular data for drug discovery. A biotech engineer trained at AgroParisTech, he spent more than 15 years in deep tech, notably helping scale Hello Tomorrow into a global platform for breakthrough science, and leading an innovation studio to its acquisition by Capgemini Invent. In 2023, he co-founded Generare to address a key bottleneck in drug discovery: access to genuinely new molecular data.

References

1. Cooper VS, Honsa E, Rowe H, et al. Experimental evolution in vivo to identify selective pressures during pneumococcal colonization. mSystems. 2020;5(3):e00352-20. doi: 10.1128/mSystems.00352-20
2. Chang X, Zhang T, Zhang W, Zhao Z, Sun J. Natural drugs as a treatment strategy for cardiovascular disease through the regulation of oxidative stress. Oxid Med Cell Longev. 2020;2020:5430407. doi: 10.1155/2020/5430407
3. Generare. Generare raises €20M after generating more novel high quality molecules in 2025 than the rest of the field combined. Published April 2, 2026. Accessed April 17, 2026.