Ring of glowing particles on a dark background

Synergy Research

Effects come from combinations,
not single molecules.

A real botanical effect is the sum of dozens of compounds acting together. Eybna models that entourage at the receptor level, the combinatorial space a single-molecule approach can never see.

OUR METHOD

The synergy thesis

A flower can contain 140+ terpenes; nature has produced over 50,000. The feeling a plant delivers, calm, lift, focus, is never a single molecule. It is a blend acting across multiple receptors at once.

Most of the industry still formulates around single “hero” compounds because that’s what’s easy to measure. It misses the interactions that actually produce the outcome. Eybna was built on the opposite premise.

Bottle with colorful swirling contents

Combinatorial screening

BNA™ explores compound combinations across 2,200+ molecules - a space too large to test by hand.

Bottle with colorful swirling contents

Combinatorial screening

BNA™ explores compound combinations across 2,200+ molecules - a space too large to test by hand.

Bottle with colorful swirling contents

Combinatorial screening

BNA™ explores compound combinations across 2,200+ molecules - a space too large to test by hand.

Automated dispensing rig with analysis screen

Receptor co-activation

We score how a blend engages multiple receptor systems together, predicting the accumulated functional effect.

Automated dispensing rig with analysis screen

Receptor co-activation

We score how a blend engages multiple receptor systems together, predicting the accumulated functional effect.

Automated dispensing rig with analysis screen

Receptor co-activation

We score how a blend engages multiple receptor systems together, predicting the accumulated functional effect.

Face mapped with glowing sensor points

Validated blends

Top candidates are confirmed in-vitro and on EEG - proof the synergy produces the intended outcome.

Face mapped with glowing sensor points

Validated blends

Top candidates are confirmed in-vitro and on EEG - proof the synergy produces the intended outcome.

Face mapped with glowing sensor points

Validated blends

Top candidates are confirmed in-vitro and on EEG - proof the synergy produces the intended outcome.

Why it matters

Why single-molecule formulation misses it.

Man sitting relaxed at home in soft daylight

Optimize one molecule and you optimize one number. Optimize a system and you can tune the whole experience: onset, intensity, the balance between two effects, the way a flavor lands first sip or puff to last

This is the difference between a flavor note that hints at an effect and an ingredient engineered to deliver it. It’s also why our 37-terpene BUUZ® formulation beat Givaudan’s Zensera™ to win Ingredient Idol 2025.

platform | layer 01

AI & In-Silico

Man sitting relaxed at home in soft daylight

Before a single drop is mixed, BNA™ models how each compound meets each receptor, screening thousands of molecules computationally, so the lab only ever tests the candidates most likely to work.

Why it matters

Modeling biology at the molecular level

In-silico is where speed comes from. Molecular docking predicts how strongly a compound binds a target receptor; molecular dynamics simulates how that interaction behaves over time. Run across 2,200+ botanical compounds and 22+ receptors, this turns a brief into a ranked shortlist in hours instead of months.

Our stack runs multiple engines for cross-validation, not a single black box:

Glowing digital render of a botanical leaf

AutoDock Vina

A molecular docking engine that predicts how small molecules bind to protein targets, rapidly scoring and ranking candidate compounds by predicted binding affinity.

AutoDock Vina

A molecular docking engine that predicts how small molecules bind to protein targets, rapidly scoring and ranking candidate compounds by predicted binding affinity.

AutoDock Vina

A molecular docking engine that predicts how small molecules bind to protein targets, rapidly scoring and ranking candidate compounds by predicted binding affinity.

GNINA

A deep learning-based docking tool that combines convolutional neural networks with traditional docking to more accurately score and rank protein-ligand poses, improving on physics-based methods alone.

GNINA

A deep learning-based docking tool that combines convolutional neural networks with traditional docking to more accurately score and rank protein-ligand poses, improving on physics-based methods alone.

GNINA

A deep learning-based docking tool that combines convolutional neural networks with traditional docking to more accurately score and rank protein-ligand poses, improving on physics-based methods alone.

GROMACS

A high-performance molecular dynamics simulation package used to model the physical movements of atoms and molecules over time, revealing how proteins and ligands behave and interact in realistic, dynamic conditions.

GROMACS

A high-performance molecular dynamics simulation package used to model the physical movements of atoms and molecules over time, revealing how proteins and ligands behave and interact in realistic, dynamic conditions.

GROMACS

A high-performance molecular dynamics simulation package used to model the physical movements of atoms and molecules over time, revealing how proteins and ligands behave and interact in realistic, dynamic conditions.

OpenMM

A flexible molecular dynamics toolkit that enables custom simulation workflows, often used to validate and refine docking predictions with detailed physics-based modeling.

OpenMM

A flexible molecular dynamics toolkit that enables custom simulation workflows, often used to validate and refine docking predictions with detailed physics-based modeling.

OpenMM

A flexible molecular dynamics toolkit that enables custom simulation workflows, often used to validate and refine docking predictions with detailed physics-based modeling.

Multi-model consensus scoring

An approach that combines predictions from multiple independent algorithms to reduce individual model bias, producing more reliable and robust rankings of candidate compounds.

Multi-model consensus scoring

An approach that combines predictions from multiple independent algorithms to reduce individual model bias, producing more reliable and robust rankings of candidate compounds.

Multi-model consensus scoring

An approach that combines predictions from multiple independent algorithms to reduce individual model bias, producing more reliable and robust rankings of candidate compounds.

Predicted, then ranked

From thousands of compounds to a shortlist.

2,200+ plant compounds modeled against 22+ human receptors, predicted, validated, and confirmed on real people.

Diagram tracing botanical source to molecular signature, olfactory interaction and brain response, ending in a predicted human state of calmness, focus, mood and stress

01

Dock

Predict binding strength between every candidate compound and the target receptor.

02

Simulate

Model the interaction over time to filter out unstable or implausible binders.

03

Rank

Score and shortlist, handing the lab a small, high-probability candidate set.

Speed as a moat

From trend signal to formula in 30 days.

Our in-silico platform turns emerging trends into market-ready ingredients within days, before traditional flavor houses can even scope the project.

Gloved hands decanting a terpene formulation into an aluminium bottle

When a new strain or flavor trend appears, we formulate against it the same week and put a sample in a customer's hands before a flavor house has scoped the brief.

Flavor houses are built around 6–18-month projects and large minimums. BNA™ is built around 30 days and 250 mL. That is structural, not a promotion.

platform | layer 02

Receptor Intelligence

Man sitting relaxed at home in soft daylight

A computational prediction is only a hypothesis until a human cell agrees. Layer 02 runs live in-vitro assays that confirm exactly how a compound engages each receptor, producing mechanism-of-action data that flavor houses do not generate.

Measured on human cells

We run functional assays on human cell lines to measure receptor response directly, not in animals, not by proxy. Three readouts tell us how, and how strongly, a compound acts:

Microscope image of a plant stem cross-section showing vascular bundles

β-arrestin recruitment

the signalling pathway behind many functional effects.

β-arrestin recruitment

the signalling pathway behind many functional effects.

β-arrestin recruitment

the signalling pathway behind many functional effects.

cAMP modulation

whether a compound activates or dampens a receptor.

cAMP modulation

whether a compound activates or dampens a receptor.

cAMP modulation

whether a compound activates or dampens a receptor.

Calcium flux

fast, sensitive confirmation of receptor engagement.

Calcium flux

fast, sensitive confirmation of receptor engagement.

Calcium flux

fast, sensitive confirmation of receptor engagement.

The result is a validated map from compound to receptor to outcome, for sleep, focus, recovery and mood, built over a decade of assays, and ours.

The receptor map

Receptors we target, and what they do.

RECEPTOR

SYSTEM

OUTCOME

CB1

CB1

Endocannabinoid

Endocannabinoid

Mood, appetite, pain modulation

Mood, appetite, pain modulation

CB2

CB2

Endocannabinoid

Endocannabinoid

Inflammation, recovery, immune balance

Inflammation, recovery, immune balance

5-HT1A

5-HT1A

Serotonergic

Serotonergic

Calm, anti-anxiety, mood lift

Calm, anti-anxiety, mood lift

GABA-A

GABA-A

Inhibitory

Inhibitory

Relaxation and sleep

Relaxation and sleep

TRPV1

TRPV1

Sensory / thermo

Sensory / thermo

Cooling, warming, pain signalling

Cooling, warming, pain signalling

A2A

A2A

Adenosine

Adenosine

Alertness and focus

Alertness and focus

α7 nAChR

α7 nAChR

Cholinergic

Cholinergic

Cognitive sharpness, without the addiction loop

Cognitive sharpness, without the addiction loop

Representative targets from a panel of 22+ human receptors. Full assay data available under NDA.

platform | layer 03

Human Brainwave Measurement

Man sitting relaxed at home in soft daylight

Anyone can claim a terpene “helps you relax.” We can show which receptor it engages, how strongly, and the human data behind it. That’s the difference between folklore and a functional claim a brand can stand behind.

It’s also what makes Eybna INSIDE™ credible: when a wellness brand needs to explain why a product works, the answer comes from this map.

A lab technician in protective gear working behind a pipette

VALIDATION & SAFETY

The right claim for the right market, by default.

RCT

Double-blind clinical

Three randomised, double-blind, placebo-controlled studies. 485 participants.

Clinical trial vial

RCT

Double-blind clinical

Three randomised, double-blind, placebo-controlled studies. 485 participants.

Clinical trial vial

Validated

FEMA & EFSA

Regulatory validation complete across food and beverage applications.

FEMA and EFSA approved stamp

Validated

FEMA & EFSA

Regulatory validation complete across food and beverage applications.

FEMA and EFSA approved stamp

100%

All Natural

Botanically derived - Never synthetic. Documented on every batch.

Botanical sprig in a vial

100%

All Natural

Botanically derived - Never synthetic. Documented on every batch.

Botanical sprig in a vial

Certifications

The documentation, in full.

Current certificates and the full compliance packet, ready to download.

GMP, Good Manufacturing Practice

ISO 9001:2015, Quality Management

FSSC 22000, Food Safety Management

Compliance Packet 2026

Compliance, built in

Shipped, compliant, certified.

BNA™ carries a regulatory layer across categories and territories, so every formula ships ready for where it’s going.

Neon-lit city street lined with cannabis and CBD retail signage

Markets

Each formula is mapped to the regulatory requirements of its destination market, helping ensure the right ingredients, claims, and documentation are in place from the start.

United States

United States

European Union

European Union

United Kingdom

United Kingdom

Japan

Japan

Australia

Australia

Categories

Compliance is tailored to the product format, accounting for the distinct safety, formulation, labeling, and claims requirements of each category.

Food

Food

Beverage

Beverage

Supplement

Supplement

Cosmetic

Cosmetic

Cannabis

Cannabis

Tobacco alternative

Tobacco alternative

Shelf packed with supplement and functional wellness products

Why it’s a moat

Globally credited.

Two researchers in cleanroom gear running a sensory evaluation

FSSC 22000, GMP, ISO 9001 and OU Kosher across three production sites. FEMA GRAS ingredients. Per-market dossiers for the US, EU, UK, Japan and Australia, generated with the formula. Global brands are raising the bar on flavor provenance; Eybna was built above it.

Woman with eyes closed, glowing aroma lines across her face

Start with a sample

Tell us the feeling you want your product to deliver. We'll formulate it in 30 days.