Flare

Flare™ is a complete computer-aided drug design (CADD) solution developed by Cresset for the design, discovery, and optimization of small molecules. Built with both computational and medicinal chemists in mind, Flare combines an intuitive graphical user interface with cutting-edge physics-based and AI-driven methods, making it a comprehensive computational toolbox for accelerating the development of novel therapeutics.

Easily deployed as a desktop application, Flare is designed to streamline molecular discovery workflows — enabling data to flow in and actionable insights to flow out. By focusing resources on the most promising molecules before laboratory experiments, teams can save time, reduce costs, and maximize the probability of success in later-stage drug development. Flare is also extensible via a Python API, allowing seamless integration with internal workflows and pipelines.

Core Capabilities

• AI/ML Tools: A suite of artificial intelligence and machine learning tools that enhance productivity, support decision-making, and simplify complex computational processes, including AI Assistants and Copilots integrated directly into the platform.
• Binding Affinity Predictions: Accurate predictions of how well molecules bind to biological targets, with a choice of methods including Free Energy Perturbation (FEP) and absolute FEP calculations.
• Compound Idea Generation: Best-in-class physics- and AI-based new molecule generation capabilities for designing novel molecular structures likely to bind a given biological target, using structural insights to guide ideation.
• Protein Modeling: Tools to create accurate models of protein structures and understand binding mechanisms, including protein-protein docking and pocket detection for uncovering druggable binding sites.
• SAR Analysis: Summarize and visualize complex Structure-Activity Relationships in both 3D and 2D to guide medicinal chemistry decisions.
• Molecular Dynamics Simulations: Study conformational changes of proteins and assess the stability of protein-ligand complexes over time.

Key Features and Methods

• Access to dozens of innovative simulation, prediction, and analysis methods delivering robust, reproducible results.
• Free Energy Perturbation (FEP) for highly accurate binding affinity predictions, including support for ligands with different charge states within the same perturbation.
• Generative AI tools for novel molecule design, introduced in Flare V11.
• Constrained protein-protein docking capabilities.
• Covalent docking features valued for their intuitive usability by medicinal chemistry teams.
• 3D-RISM water analysis for exploring binding site hydration and informing drug design.
• Protein interaction potential visualization to highlight common features across target families and drive ligand design.
• QSAR modeling (2D and 3D) for prioritization of new molecule designs.
• Scaffold hopping and bioisosteric replacement studies to identify structurally diverse active compounds.
• Pocket detection to explore the structural landscape of target proteins and uncover potential binding sites.
• Macrocycle design and prioritization support for computationally challenging compound classes.
• Detached calculations enabling background processing to improve workflow efficiency.

Workflow and Usability

• Designed by discovery chemists to minimize the learning curve through an intuitive graphical user interface.
• Supports end-to-end small molecule discovery workflows, from initial compound ideation through optimization and prioritization.
• Flare Python API enables easy integration with internal informatics and computational workflows.
• AI Assistants include a Chat Model for documentation-aware guidance on GUI features and a Code Model for scripting support.
• Empowers medicinal chemistry teams to make data-driven decisions without requiring deep computational expertise.

Applications and Scientific Use Cases

• Identification and optimization of inhibitors for challenging targets, including METTL3, TYK2, and SARS-CoV-2 Mpro cysteine protease.
• In silico prioritization of macrocyclic designs for synthesis, reducing synthetic effort on lower-probability candidates.
• Free Energy Perturbation studies to accurately predict binding affinities of new ligand suggestions.
• Exploration of new binding opportunities through pocket detection and protein interaction potential analysis.
• QSAR-driven prioritization of compound libraries for viral and oncology targets.

Flare is delivered as a desktop application and is regularly updated with new scientific features and enhancements — recent releases include Flare V10 and V11, introducing absolute FEP calculations, generative AI tools, protein-protein docking, and expanded AI Copilot functionality. The platform is trusted by computational and medicinal chemists at leading pharmaceutical, biotech, and academic organizations worldwide.