Applications

0.3

kcal/mol

Binding accuracy vs experiment

We predict protein–ligand binding free energies from first principles, then check them against the lab. Across CDK2, MCL1 and Thrombin the model reaches chemical accuracy, and on the hardest case — CDK2 — our neural-network term lifts the correlation with measured affinities to 0.88. That is the line between a model that ranks molecules and one that tells you which to make next.

Binding and lead optimization for proteins and ligands

Potential applications

Lead optimisation — push affinity from micro toward nano/picomolar
Atom- and functional-group guidance for medicinal chemistry
Selectivity and off-target / toxicity-liability screening
Binding-pocket and cryptic-site identification
Resistance-mutation and protein-variant effects on binding
Covalent, metal-binding and charged ligands beyond classical force fields

Drug discovery

Solvation

Solubility and partition of liquid mixtures

Aqueous solubility and dissolution behaviour
Partition and distribution coefficients (logP / logD)
Membrane permeability and ADME / PK properties
Formulation, co-solvent and excipient screening
pKa- and protonation-dependent solubility

Potential applications

0.2

kcal/mol

Solvation accuracy vs experiment

How a molecule dissolves and partitions underlies everything from formulation to dosing. We compute solvation free energies to chemical accuracy — and partition coefficients to within 0.22 log units — built entirely from quantum mechanics, with zero fitting to experimental data. The model has never seen the answer, and still gets it right.

Energy

Electrolyte and electrode R&D for batteries

Validated internally — data under partner agreement, publication in progress

Battery performance comes down to how ions move through an electrolyte — charged, strongly-interacting chemistry that classical models struggle to capture. Our engine handles exactly these interactions from first principles, the same physics it reproduces against quantum mechanics. We bring it to electrolyte formulation, interphase chemistry and cathode degradation — the building blocks of safer, faster-charging cells.

Potential applications

Electrolyte formulation and additive design (Li⁺ and beyond)
Ion solvation, transport and conductivity
Electrode–electrolyte interphase (SEI) and degradation pathways
Fast-charging and thermal-stability optimisation
Next-generation chemistries — solid-state, sodium, multivalent

Biochemistry

Reaction energetics in enzyme modeling

Reaction mechanisms and transition-state energetics
Catalytic pathways and intermediate stability
Enzyme engineering and active-site mutation effects
Cofactor and metal-centre chemistry
pH and protonation-state effects in catalysis

Potential applications

Validated internally — data under partner agreement, publication in progress

Nature's catalysts: tracking energies along a reaction path in full, explicit solvent. Our engine is built to clear. With the energetics of water and ions settled, we turn the model to catalytic and enzymatic mechanism: how bonds form and break, and how a single mutation reshapes activity.

Materials

Process and materials simulation for metals

Potential applications

Mineral processing and ore-enrichment chemistry
Surface and interface adsorption / separation
Corrosion and materials-degradation pathways
Catalyst and surface reactivity
A first-principles platform extensible across materials problems

Validated internally — data under partner agreement, publication in progress

The physics that governs molecules governs materials. We extend the same first-principles engine to industrial materials and process chemistry — beginning with an ore-enrichment study — as a platform that generalises across problems rather than a model rebuilt for each one.