Orion Small Molecule Discovery Suite

The Orion Small Molecule Discovery Suite, developed by OpenEye Scientific (now part of Cadence), is a cloud-native molecular design platform that provides complete solutions for all stages of small molecule drug discovery — from target exploration and hit identification through to lead optimization. Rooted in trusted OpenEye science and optimized for performance, the suite delivers a comprehensive range of ligand-based and structure-based modeling workflows accessible entirely from a web browser, enabling scientists to build models, run calculations, analyze results, and collaborate with colleagues in real time.

Designed for medicinal chemists, computational scientists, and drug discovery teams of all sizes, the Orion Small Molecule Discovery Suite combines ease of use with massive throughput and cloud scalability. Automated guided workflows — called Floes — eliminate the need for complex infrastructure setup, while curated databases and integrated cheminformatics tools allow users to work across the full discovery pipeline without visiting multiple platforms or preparing external databases.

Target Exploration Capabilities

• Automated Protein Preparation and Structural Quality Assessment
• Protein Binding Site Similarity Searching via SiteHopper
• Cryptic Pocket Detection with ligandability prediction and optimized Dynamic Probe Binding Analysis
• Structural Biology Floes using cryo-EM data to generate protein conformer ensembles with Weighted Ensemble Molecular Dynamics (WEMD)
• Sequence-to-Structure prediction using AI Fold, enhanced with Boltz-2-based improvements for more accurate folding models and better conformational state sampling

Hit Identification Capabilities

• Ligand-based and structure-based virtual screening workflows
• Ultra-large 3D docking with Gigadock™ and Gigadock Warp (AI-enabled, supporting billions of compounds with up to 40% compute cost savings)
• Very fast 3D shape similarity searching with FastROCS™
• Extreme-scale 3D shape similarity searching in synthon space with ROCS XTM
• Large-scale library enumeration and large-scale compound clustering (scaffold, fingerprint, and shape-based clustering across millions of molecules)
• 2D and 3D similarity searching via Molecule Search
• Shape alignment and scoring with ROCS®
• Electrostatic similarity searching
• Conformer generation and sampling
• Ligand-guided pose prediction in binding sites
• Molecular docking and scoring
• Solvent mapping and thermodynamic calculations

Lead Optimization Capabilities

• Generative library enumeration and large-scale compound clustering
• Fragment replacement and scaffold hopping
• 3D-QSAR modeling with refined methods and a weighted consensus model for improved potency predictions
• Electrostatic similarity lead hopping
• Off-target prediction and property filters and assessment
• Binding affinity and free energy calculations, including Non-Equilibrium Switching Star Maps with automated hub selection
• Docking, ligand posing, and induced fit posing
• Membrane permeability and mechanism analysis
• Short trajectory MD and analysis
• Full molecular dynamics with selectable force fields, including support for user-generated bespoke force fields
• Interactive Edge Mapper for seamlessly adding or removing edges and generating star maps for free energy calculations

Cheminformatics, Data Preparation, and Analysis

• Tautomer and charge assignments
• Data and file conversions
• Automated preparation for large databases
• Dataset similarity and clustering
• Ionization state enumeration with a multistate pKa model
• pKa Predictor Floes for estimating dominant ionization states to improve solubility, permeability, and binding affinity modeling

Quantum Calculations

• Psi4-based energy calculations and geometry optimization
• Support for a variety of solvent models
• Torsion profile analysis
• Conformer ensemble generation
• Gaussian Module available with additional licensing

Machine Learning

• Build and train machine learning models, including graph convolutional neural network (GCNN) models learned directly from molecular graphs
• Predict, explain, and visualize predicted physical properties
• Built-in analysis to highlight data-to-parameter balance and flag potential overfitting risks for more robust predictive models

Curated Database Access

• Access to more than 24 billion stereoenumerated commercially available compounds
• More than 100,000 prepared protein structures
• More than 2 million protein binding sites (known and putative), searchable in minutes via SiteHopper

Key Reasons to Choose Orion Small Molecule Discovery Suite

• Robust science in easy-to-use workflows: Leverage trusted OpenEye science through ready-to-use Floes without requiring deep computational expertise
• Convenience: Access all data preparation, modeling, and analysis workflows from a unified web environment — no hardware maintenance or data backup required
• Streamlined experience: Sketch, copy, and edit molecules or queries and perform 2D and 3D searches of commercially available compounds from multiple vendors, all in one place
• Time and cost savings: Workflows are optimized for speed, cloud performance, and cost efficiency
• No restrictions: No licensing restrictions on the number of modeling tasks performed or the size of calculations, enabling work on the most challenging therapeutic projects

The Orion Small Molecule Discovery Suite is delivered as a web-based cloud platform requiring only a browser, but OpenEye also supports desktop and Linux applications for local hardware preferences (Linux, Windows, or macOS), development toolkits for building custom modeling and cheminformatics solutions, and consulting services backed by decades of pharmaceutical research expertise. The suite is continuously updated, with recent releases introducing Boltz-2-enhanced AI Fold, cryo-EM-guided Structural Biology Floes, GCNN-based ML model building, new pKa prediction capabilities, and large-scale compound clustering Floes.