FUn

FUn provides interactive 3D visualization for record-level inspection and exploration of large scientific datasets.

Key Features:

• Interactive 3D Visualization: Supports interactive 3D visualization of datasets with more than 100,000 data points.
• Scalability: Leverages modern graphical processing units (GPUs) to render millions of data points efficiently on consumer hardware including laptops, tablets, and mobile phones.
• Record-level Inspection: Enables detailed inspection of individual records within large datasets, beyond summary statistics or reduced representations.
• Client-Server Architecture: Comprises Faerun (client) for rendering and Underdark (server) for data management and serving.
• Preprocessing Toolchain: Includes Lore.js and a comprehensive data preprocessing toolchain for preparing large datasets for visualization.

Scientific Applications:

• Cheminformatics and Drug Discovery: Applied to large-scale chemical patent data such as SureChEMBL (over 17 million chemical compounds), supporting visualization of large compound libraries and complex molecular data.

Methodology:

FUn uses a client-server architecture where Faerun handles rendering of interactive 3D visualizations optimized for modern GPUs and Underdark performs data preprocessing and serves datasets; the framework incorporates Lore.js and a dedicated data preprocessing toolchain.