BioPharmics

BioPharmics is Optibrium's integrated ligand-based and structure-based drug design platform, built to deliver fast, accurate, and robust 3D computational modelling across a wide range of molecular complexity — from small molecules through to large peptidic macrocycles. The platform is designed for medicinal chemists, computational chemists, and drug discovery teams who need reliable virtual screening, binding affinity prediction, and conformational analysis without sacrificing speed or scientific rigour.

BioPharmics combines several industry-leading methods — ForceGen, xGen, Surflex™-Dock, eSim, and QuanSA — into a cohesive platform that supports every stage of 3D molecular design. Its capabilities have been validated in peer-reviewed publications and demonstrated in real-world collaborations, including a landmark case study with Bristol Myers Squibb that achieved a potential 90% reduction in synthesis and testing effort during macrocycle optimisation.

3D Chemical Toolkit Operations

• Generate fast, accurate, and template-free conformational ensembles, even for complex macrocycles, using the patented ForceGen method
• Produce 3D structures from 2D inputs, including chirality enumeration and heuristic protonation at physiological pH
• Build large-scale conformer databases suitable for virtual screening campaigns
• Apply NMR constraints to increase conformational search efficiency

Docking and Structure-Based Design with Surflex™-Dock

• Automate protein preparation and alignment to reduce inappropriate bias and streamline workflows
• Access top-tier docking enrichment and highly accurate pose prediction through extensively validated, fully automated procedures
• Model conformational heterogeneity of bound ligands using real-space fitting of ligands into X-ray electron density via xGen
• Use ensembles of protein conformations to account for binding pocket variability and capture relevant interactions more realistically
• Perform highly accurate non-cognate ligand docking to predict poses for future potential ligands using known prior bound poses with eSim
• Access Surflex™-Dock predictions through an intuitive, highly visual PyMOL GUI
• Refine conformational ensembles using X-ray electron density maps with xGen, producing more accurate, lower-strain models of ligand binding — accessible to non-crystallographers and within crystallographic workflows

Ligand-Based Modelling with eSim

• Screen millions of compounds with industry-leading enrichment using eSim's 3D virtual screening capabilities
• Predict relative bound poses of structurally diverse ligands through multiple-ligand alignment
• Create surface representations of molecules based on shape, electrostatic field, and hydrogen-bond preferences, reflecting a protein's perspective of a ligand
• Achieve significantly higher enrichment in virtual screens compared to leading competitors
• Inform 3D design with industry-leading pose prediction
• Access eSim as part of the StarDrop drug discovery platform for fully integrated, intuitive drug discovery workflows

Binding Affinity Prediction with QuanSA

• Predict ligand binding affinity and pose using multiple-instance machine learning through the QuanSA (Quantitative Surface-field Analysis) method
• Access physically motivated, accurate, and scaffold-independent models that achieve equivalent accuracy to market-leading methods such as FEP+
• Predict binding affinity without requiring protein structural information
• Visualise the key interactions driving molecular affinity through the QuanSA plugin for PyMOL, guiding the design of more potent compounds
• Build accurate affinity models even with only a small number of known actives
• Identify novel active compounds with prediction confidence metrics

Demonstrated Impact: Macrocycle Optimisation Case Study

• In collaboration with Bristol Myers Squibb, BioPharmics was applied to the PD1/PD-L1 system to optimise the macrocyclic lead Pep-01 to the clinical candidate BMT-174900
• Conventional approaches would require thousands to hundreds of thousands of analogues; BioPharmics accurately prioritised a focused set of just 72 analogues
• The clinical candidate ranked in the top 10% of compounds, demonstrating a potential 90% reduction in synthesis and testing effort
• The approach combined NMR-validated bioactive conformations, guided conformational search, structure-based docking, and explicit modelling of ligand strain

Scientific Publications and Peer-Reviewed Research

• Active learning applied to scaffold replacement for identifying mimics of macrocyclic natural products, maintaining potency, biodistribution, metabolic stability, and synthesis scalability
• A distributional model of bound ligand conformational strain, covering small molecules through to large peptidic macrocycles
• Structure- and ligand-based virtual screening benchmarked on DUD-E+, exploring the impact of single versus ensemble protein pocket and ligand representations
• A wide range of additional peer-reviewed publications covering ForceGen, xGen, Surflex™-Dock, eSim, and QuanSA methods

BioPharmics is delivered by Optibrium and is verified for use by customers across the life sciences industry. The platform integrates with PyMOL for visualisation and with Optibrium's StarDrop drug discovery platform, enabling 3D ligand-based design as part of a broader, integrative drug discovery workflow. Virtual screening collections, including Enamine's commercially available screening library, are available for use within compatible StarDrop modules.