Registered & recognized
Five Angstrom designs and runs the AI and computational methods behind precision analytics, molecular research, and secure infrastructure — for hospitals, institutions, and research groups who need answers, not just software.
We are, first and foremost, an AI & Scientific Computing company — the models, simulations, and infrastructure in this site all come from the same research group, not separate teams bolted together.
Generative AI and molecular modelling work, run end-to-end by our own research group for partners who need the science done, not just the software.
Predictive models, molecular dynamics, and statistical rigor applied to real clinical and institutional data — built to hold up under scrutiny.
Encryption designed to stay secure even against future quantum computers, built for partners handling sensitive institutional data today.
Five active research threads — explained in plain language first, with the method named alongside it for anyone who wants the technical detail. Read the full research page →
PETase is a natural enzyme that breaks down PET plastic — the kind used in bottles. On its own, it's slow. We use generative AI models to design new, more effective versions of it: rather than testing enzyme variants one at a time in a lab, our models learn the shape and chemistry of enzymes that work, and generate new candidate structures likely to work better.
In plain terms: think of it as AI sketching thousands of possible tools shaped to do one job — cutting plastic apart faster — so scientists only need to test the most promising few in the lab.
A peptide is a short chain of amino acids — smaller and simpler than a full protein. Many diseases involve a specific molecule that a peptide could attach to and block or signal. Instead of testing thousands of peptide sequences by hand, we generate candidates computationally, aimed at a chosen target shape.
In plain terms: it's like designing a key for a specific lock — except the AI proposes hundreds of candidate keys at once, ranked by how well they're likely to fit, before anyone cuts real metal.
Drug discovery usually starts from an enormous space of chemically possible molecules — far too many to synthesize and test. Our generative models learn what makes a molecule likely to work against a given target, and produce a much smaller set of candidates worth actually making.
In plain terms: instead of searching a library with millions of books one at a time, the AI reads the whole library at once and hands you the dozen that are actually worth opening.
Molecules like proteins and peptides don't hold one fixed shape — they fold into many possible 3D arrangements, called conformations. Simulating all the plausible ones is expensive if you sample randomly. Digital nets are a mathematically structured way of choosing sample points that cover the possibility space more evenly and efficiently than random guessing.
In plain terms: imagine mapping every room in a huge building — random sampling means wandering corridors hoping to stumble onto each room, while our method walks a planned route that covers every wing with far fewer steps. We've released this algorithm as open source for other researchers to use.
Two active collaborations putting this research to work in clinical settings. Read the full case studies →
Building an AI model to help predict liver tumors from medical imaging — aiming to give clinicians an earlier, more consistent second read alongside their own diagnosis.
The work applies our modelling and analytics methods to real patient imaging data, under a formal research collaboration with the hospital.
AI-based liver tumor predictionStudying biomarkers linked to dental disease, combined with molecular dynamics simulations — a computational technique that models how molecules move and interact over time.
The goal is to connect measurable biological signals with the underlying molecular behaviour driving dental disease.
Biomarkers & MD simulationMost encryption in use today could eventually be broken by a sufficiently powerful quantum computer. Quantum-secure encryption is designed using mathematical problems that stay hard even for quantum machines — so data protected today stays protected years from now.
Data encrypted today can be recorded and stored by an attacker, then decrypted later once quantum computing catches up. For institutions holding long-lived sensitive records, that risk exists now, even though the technology to exploit it doesn't yet.
Five Angstrom designs and implements quantum-secure encryption for partner platforms handling sensitive institutional data — protecting records against both today's and tomorrow's methods of attack.
Whether it's a molecule to design, data to model, or infrastructure to secure — talk to the research group directly.