XMolGen

XMolGen is an integrated AI-driven molecular design platform developed by XtalPi (晶泰科技), purpose-built to accelerate drug discovery by combining artificial intelligence and big data technologies. The platform is designed for medicinal chemists and drug discovery teams seeking to overcome the limitations of traditional high-throughput screening and synthetic chemistry approaches, including high R&D costs, constrained chemical space exploration, and dependence on known compound libraries.

By integrating three core generation engines — R-group replacement (R基替换), scaffold replacement (骨架重构), and de novo generation (全新生成) — XMolGen enables high-throughput, algorithm-driven molecular generation that systematically expands the boundaries of conventional chemical space exploration. The platform covers molecular generation, molecular screening, molecular docking, and intelligent retrieval of commercial compound libraries within a single unified environment.

Core Generation Engines

• R-group Replacement (R基替换): Ligand-based generation that substitutes functional groups around a fixed core scaffold to produce diverse analogues.
• Scaffold Replacement (骨架重构): Replaces the central scaffold of a molecule while retaining key pharmacophoric features, enabling structural novelty.
• De Novo Generation (全新生成): Creates entirely new molecular structures from scratch, empowering medicinal chemists to explore previously uncharted chemical space.

Key Capabilities and Features

• Multi-scenario support: Accommodates both ligand-based methods (R-group and scaffold replacement) and receptor-based methods (binding pocket-guided molecular generation), making it adaptable to diverse drug discovery contexts.
• Mature AI model integration: Incorporates a range of algorithms including traditional approaches such as retrosynthesis, fragment assembly, and bioisostere replacement, alongside innovative AI-driven algorithms for enhanced molecular design.
• High-throughput generation: Algorithm-driven workflows enable the rapid generation of large numbers of structurally diverse and novel candidate molecules.
• Physicochemical and structural filtering: Supports multiple filtering conditions based on physicochemical properties and structural rationality to ensure generated molecules meet drug-likeness criteria.
• User-friendly interface: Features a clean, intuitive layout with straightforward operational workflows, significantly lowering the learning curve for new users.
• Validated in real projects: XtalPi has applied XMolGen across real-world drug development projects, demonstrating the platform's ability to generate novel, synthetically accessible, and structurally sound molecules.

Application Scenarios

• De novo molecular generation: Enables the creation of novel compounds from zero, with support for physicochemical property filters and structural rationality checks to facilitate innovative drug discovery.
• Compound library generation: Presents generated molecular results in a clear and concise manner, displaying corresponding reaction pathways and building block information for each molecule.
• Virtual screening: Supports batch docking of generated molecules, enabling efficient evaluation of large compound sets to identify promising research candidates.

Workflow and Integration

1. Select a generation mode suited to the project context: ligand-based (R-group or scaffold replacement) or receptor-based (binding pocket generation).
2. Apply physicochemical and structural filters to refine the generated molecular pool.
3. Review compound library outputs with associated reaction routes and building block details.
4. Perform batch virtual screening and docking to prioritise high-potential candidates for further development.

XMolGen has been validated through real drug discovery projects at XtalPi, including lead compound optimisation targeting flexible allosteric protein binding sites. The platform is available via a free trial application, and its design reflects XtalPi's broader commitment to integrating computational intelligence into every stage of pharmaceutical research and development.