Biologix · Physics-grounded polymer discovery

Designing the polymers that keep biologics stable without a cold chain.

Biologix is an agentic discovery platform that reasons over chemistry the way a scientist does — proposing candidate polymers, testing them against real physics, and closing the synthesis route before it moves on.

Screening loop Human checkpoint every iteration
TARGET n REPEAT UNIT
Stage Molecular screening

01 — The problem

A medicine that only works if it stays cold isn't reaching everyone.

Insulin and the biologics that follow it must be kept refrigerated. Where refrigeration is unreliable or absent, the medicine degrades before it can help — a supply problem hiding inside a materials problem.

A protective polymer matrix could stabilize these proteins at ambient temperature. But the space of possible polymers is astronomically large, and no experimental campaign can screen more than a sliver of it — let alone check whether each candidate can actually be made.

500M+

people worldwide depend on insulin

2–8°C

the cold chain today's supply is trapped inside

02 — The shift

Surrogate models guess at chemistry. Biologix reasons over it — and checks its answers against real physics.

PROPOSE MEMORY JUDGE

A language model as scientist

An LLM proposes hypotheses and reads the literature — acting as the acquisition function that decides what to try next.

A persistent discovery world

Every hypothesis, simulation result, and literature claim accumulates into a shared memory that sharpens each iteration.

Physics as the judge

Explicit molecular simulation rejects infeasible packings and implausible structures before they can steer the search.

03 — How it works

A closed loop that runs end to end, with a scientist in it.

Candidate generation, molecular simulation, retrosynthesis, and safety screening operate as one loop — grounded in physics rather than interpolation, with a human checkpoint on every iteration.

01

Literature mining

Reads the field to seed and constrain the search.

02

Candidate generation

Proposes and mutates polymer repeat units in PSMILES.

03

Molecular screening

Scores each candidate against the target with OpenMM & Packmol.

04

Retrosynthesis

Closes a viable synthesis route before the candidate advances.

05

Safety & compliance

Screens ADMET and excipient rules on every iteration.

04 — Capabilities

Point it at a biologic. It handles the rest.

Any target

By name or PDB ID

Resolve an arbitrary biologic and the same workflow runs unchanged — insulin and adalimumab shared one pipeline.

Polymer space

Repeat-unit search

Proposes and mutates candidate repeat units across the discrete PSMILES space under a fixed evaluation budget.

Physics scoring

Simulation as oracle

OpenMM and Packmol matrix screening scores each candidate against the real target structure, not a learned surface.

Synthesis-aware

Routes, not just structures

Agent-backed retrosynthesis and a four-tier precursor registry close a viable route before a candidate advances.

Safety by default

Screened every iteration

ADMET and excipient-compliance checks run inline on every loop — recorded as audit artifacts, not bolted on at the end.

Runs anywhere

Commodity hardware

The science stage is CPU-bound, compatible with local and edge LLM deployments — no specialized cluster required.

Three ways to search the polymer space. Only one reasons and verifies.

Wet-lab screening

Ground truth, impossible scale

The only true measurement — but no campaign can synthesize and test more than a sliver of a combinatorial space.

Surrogate ML

Fast, but guessing

Learned predictors are quick, yet they interpolate — they can't reject sterically infeasible packings or implausible structures.

Biologix

Reasons, simulates, closes the route

An LLM decides what to try; explicit physics judges it; retrosynthesis and safety gate it — every iteration grounded, not interpolated.

05 — The evidence

One workflow. Two proteins. It didn't need to change to find both.

Independent campaigns on insulin and adalimumab ran the exact same platform, unchanged — each converging on a synthesizable candidate polymer and outperforming reinforcement-learning and Bayesian-optimization baselines under matched budgets.

Campaign A · Insulin

Converged on a carbonate-lactide polymer

Lactate comonomers added the hydrogen-bonding the protein needed — a candidate the model reasoned toward and physics confirmed.

Campaign B · Adalimumab

Found a simple amide-ketone backbone

Bulky designs failed to pack; the platform learned that within the session and settled on a linear backbone that both packed and scored best.

The loop that stabilized insulin could hold for any biologic where a screening oracle exists. We're building toward a world where a medicine's reach is no longer decided by whether it can be kept cold.

Algonix AI

Press release Partnership · August 2026

Algonix AI partners with NovoMCP to bring Biologix to the open computational chemistry engine.

Edinburgh, Scotland — Algonix AI and NovoMCP have entered a partnership under which NovoMCP will use Biologix, Algonix's agentic platform for polymer-excipient discovery, within its own computational chemistry ecosystem.

NovoMCP operates an open engine that AI agents call through the Model Context Protocol. Through the partnership, its work gains Biologix's physics-grounded, synthesis-aware polymer discovery — the same workflow Algonix used to find candidate stabilizers for insulin and adalimumab.

Read the announcement →

“Biologix was built to run anywhere a screening oracle exists. Seeing it put to work inside NovoMCP's engine is exactly the reach we designed for.”

Founder, Algonix AI

Questions

For the researchers reading closely.

How does it avoid hallucinating chemistry?

Physics checks reject infeasible packings and name–structure mismatches before they can steer the next iteration.

What is the screening oracle?

OpenMM Packmol-matrix screening, evaluating interaction energy between candidate polymers and the specified protein structure.

Does it check synthesizability?

Yes — agent-backed retrosynthesis closes a route, seeded from a four-tier precursor registry, before a candidate advances.

Which targets have you run end to end?

Insulin (PDB 4F1C) and adalimumab (PDB 3WD5), each over five human-in-the-loop iterations of the same unchanged workflow.

What hardware does it need?

The science stage is CPU-bound and runs on commodity hardware, compatible with local and edge LLM deployments.

Is a run reproducible and auditable?

Every iteration logs hypotheses, simulation outcomes, ADMET, and compliance alongside the structure files it produces.

06 — Research

The platform is documented in the open. Read the science behind it.

Put Biologix to work on your biologic.

Bring us a target and a screening oracle. We'll run the same loop that found candidates for insulin and adalimumab.

Request a demo →