MolParser is a novel end-to-end system that automatically extracts and recognizes molecular structures from chemistry literature, addressing a critical gap in scientific knowledge processing.

• Enables precise extraction of chemical structures from documents
• Tackles the challenge of complex Markush structures in real-world publications
• Creates new opportunities for large-scale literature mining in chemistry
• Enhances the application of LLMs in scientific domains

This technology has significant implications for pharmaceutical research, enabling faster drug discovery through automated extraction of molecular information from the rapidly growing body of chemistry publications and patents.

MolParser: End-to-end Visual Recognition of Molecule Structures in the Wild