Biosimulytics and the EIC Accelerator: Predicting Drug Crystals

Biosimulytics, a spin-out of University College Dublin formed in 2019, predicts the crystal forms, or polymorphs, that a drug molecule can take, using physics-based simulation sped up by machine learning. An unexpected polymorph can change how a medicine dissolves or how stable it is, and has caused product recalls and patent disputes. The company was selected in the EIC Accelerator's March 2023 cut-off and received an EU grant of €2.50 million for BioSim M2M, "Molecules to Medicine", a €3.57 million project that ran from September 2023 to February 2026, according to CORDIS. The goal was to make crystal structure prediction affordable for every drug developer as a cloud-based solution, cutting a prediction cycle from three months to three weeks or less.

The project has ended, but CORDIS holds only its first periodic report, to August 2024. It says the tool moved from alpha to beta and, on test molecules from the Cambridge Crystallographic Data Centre's blind test, placed the experimentally known structure within its top 10 predictions in under three weeks, according to the CORDIS reporting page. The report also names a scale-up partnership with Viridien, a French high-performance computing company. No final results have been published.

Where Biosimulytics Stands Now

The company's website offers its CHEMINA platform through a "Fast CSP", a rapid solid-form risk assessment that it says delivers results "in days", and asks clients to contact the team rather than sign up online. It lists a core team led by chief executive and co-founder Peter F Doyle, with a chief technology officer and several computational chemists. In October 2025 Biosimulytics joined QUBIC, a quantum computing project led by Equal1 that won €13.7 million from Ireland's Disruptive Technologies Innovation Fund, according to UCD. Its share is not stated. Irish company accounts are not freely accessible, so revenue and losses are not public here.

The Technology: Mapping a Molecule's Solid Forms

Crystal structure prediction generates the ways a molecule can pack into a crystal and ranks them by stability. It is computationally heavy, because accurate energies need quantum-level calculations. Biosimulytics trains a neural network on such calculations, with active learning that adds data on the fly, to reach similar accuracy faster. The company's pitch is speed and cost, so that smaller biotechs, which it says hold 80% of the drug pipeline, can use it.

What the EIC Award Promised, and What the Record Shows

The award promised an affordable, scalable, cloud-based prediction tool for all drug developers. At the halfway point there was a beta version with promising results on blind-test molecules, new hires including a chief technology officer, and a computing partner. Whether a cloud product launched, how it performed on customer molecules and how many companies use it are not public.

1. Neural network modelDeliveredTasks completed, first report.
2. Beta CSP toolDeliveredTop 10 on test molecules.
3. Cloud product for allPartialSold as a service.
4. Final resultsMissingNo final report.

The Money

The EU paid 70% of the project. Before the EIC award, Biosimulytics raised a €595,000 seed round in 2021 with Enterprise Ireland, according to Silicon Republic. No later private round has been announced. On disclosed figures, the EIC grant is the largest source of funding the company has had.

What Cannot Be Checked

Why Europe Should Care

Solid-form problems can delay a new drug by years, the company says, and small biotechs, which develop most new molecules, rarely have the tools to spot them early. A cheaper prediction service could help. Biosimulytics has the science and a beta tool; the public record does not yet show a business that is reaching many customers.

The Verdict

Biosimulytics and the EIC have not been asked for comment for this article. It is based entirely on public records and published statements, linked throughout.