BARon
Protein-ligand free energy calculation method technology
Increases drug design success rates through precise free energy prediction.
CADD Solution Optimized for
All Drug Targets
BARon is a next-generation platform for predicting drug-target binding affinity that integrates AI and MD, moving beyond the limitations of conventional docking-centric approaches. This innovative binding free energy technology is validated by high correlation between its predicted data and experimental data across diverse protein targets.
BARon ensures highly accurate binding affinity prediction through alchemical analysis based on the BAR (Bennett Acceptance Ratio) algorithm.
Principles of the BAR Algorithm
BARon precisely calculates the protein-ligand binding free energy ΔG = ΔH - TΔS using the Bennett Acceptance Ratio (BAR) method.
Core Technical Advantages
Simultaneous Calculation Method
Calculates forward and reverse reactions simultaneously, significantly reducing computation time.
Extraction of Non-bonded Energy Only
Maximizes efficiency by selectively calculating only the λCoulomb and λVdwaals interactions.
Explicit Solvent Model
Recreates a realistic physical environment by using an explicit solvent model that includes actual water molecules.
BARon-Based Binding Affinity Prediction Process
BARon utilizes an efficient equilibration protocol to predict binding affinity within approximately half a day (based on GPCR standards).
Equilibration Phase
NTSR1-NTS Complex Equilibration
MD Simulation
100ns Molecular Dynamics Calculation
BAR Calculation
Free Energy Analysis per System
Validated Accuracy and Efficiency of BARon
- ~5x improvement in initial hit screening success rate
- Secured success in In-silico to In-vitro connection through multiple PoCs
- Published in international journals including Chemical Science (2025)
The BARon technology has demonstrated reproducibility and scalability by screening hundreds of compounds and deriving significant Hits through multiple government-backed projects and corporate PoC (Proof of Concept) validation projects.
Case Studies on BARon-Based Drug-Target
Binding Affinity Research
The accuracy of BARon is validated by its high correlation with experimental data across multiple GPCR systems.
Selective Antagonists for A1AR and A3AR
Key Findings
- Validation of the in-silico approach for ligand selectivity of A1AR and A3AR
- A1AR generally showed higher calculated-to-experimental ratios, while A3AR showed lower values in both experimental and calculated data.
- The most crucial factor influencing antagonist dissociation is the salt bridge between E169ECL2:H264ECL3.
- The importance of this salt bridge is based on the ligand residence time at the binding site, which is also related to the energy between the ligand and the target protein at the binding site.
Four Activation States of Adenosine A2AR
Key Achievements
AMX-BAR accurately predicted the experimental binding forces of agonists reacting to the Inactive (R), active-intermediate (R'), and fully active structure (R*).
Successfully reproduced the differences in binding affinity shown by the same agonist in different activation states of the GPCR using molecular dynamics-based BAR calculation.
Prediction of β1AR Active/Inactive States
Methodological Innovation
Proposed an efficient simulation protocol, reconstituting the AMX-BAR (Bennett Acceptance Ratio) method, to resolve the issue of insufficient sampling that causes discrepancies with experimental studies.
Validation Through Over 10 Years of GPCR Research
(Nat. Chem. Biol. 2018)