Sightera raises €3M to train drug-discovery AI on patient biology

University of Antwerp spinout Sightera Biosciences has raised a €3 million pre-seed round to build a drug-discovery platform trained on proprietary biological data from patient-derived samples.

What happened

Entourage, Anacura and QBIC led the financing. Sightera is initially focusing on oncology and fibrosis and plans to move its lead molecular-glue programme toward selection of a preclinical candidate.

The company uses patient-derived samples and organoids to generate experimental data, then applies AI to identify and optimise potential drug candidates. That differs from platforms trained mainly on public datasets or computational representations without an equivalent proprietary biological layer.

Sightera remains early-stage. The funding supports platform development and preclinical work rather than a clinically validated medicine.

Why it matters

AI drug discovery can generate promising molecular designs quickly, but many candidates fail because computational predictions do not capture the complexity of human biology. Starting with patient-derived responses may help Sightera filter ideas against more realistic biological systems earlier in the process.

Its molecular-glue focus is also notable because these compounds can influence disease-related proteins that are difficult to target with conventional drugs.

The bigger picture

The most defensible AI-biotech companies may be those that own unique experimental datasets rather than relying only on similar models and public information. Sightera is trying to build that feedback loop between wet-lab biology and computation.

The challenge is proving that its approach produces better molecules, not just more data. Success will depend on reproducible experiments, preclinical efficacy, safety and eventually clinical results. At this stage, the round funds a scientific thesis—not a proven treatment platform.