BioChemInsight: An Open-Source Toolkit for Automated Identification and Recognition of Optical Chemical Structures and Activity Data in Scientific Publications

2025-04-12

AI Summary

Current optical chemical structure recognition (OCSR) tools lack autonomous capability to link molecular structure images with bioactivity data, severely hindering structure–activity relationship (SAR) modeling and drug discovery. To address this, we propose the first end-to-end closed-loop framework integrating vision-language models (Qwen2.5-VL-32B, Gemini 2.0 Flash) with cheminformatics tools (DECIMER Segmentation, MolVec, PaddleOCR), enabling structure recognition, compound identifier mapping, semantic alignment of bioactivity values, and unit standardization. Our method achieves 95% accuracy in table-based structural recognition on patent data, 78–80% on non-tabular literature figures, >99% identifier matching rate, and 92.2–97.4% accuracy in bioactivity data extraction. Preprocessing time is reduced from weeks to hours. This work systematically overcomes the longstanding bottleneck in automated SAR data curation—marking the first scalable, integrated solution for end-to-end SAR knowledge extraction from heterogeneous scientific sources.

Technology Category

• Computer Vision: Medical and Biological Imaging
• Machine Learning: Structured Learning
• Data Mining & Knowledge Management: Linked Open Data, Knowledge Graphs & KB Completion

Application Category

• Search and Retrieval-Augmented AI: Web query analysis, representation and understanding
• Semantics and Knowledge: Scalable techniques for the creation, curation, publication, maintenance, and consumption of large, Web-based, structured, reusable, knowledge graphs and ontologies
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs

Abstract

Automated extraction of chemical structures and their bioactivity data is crucial for accelerating drug discovery and enabling data-driven pharmaceutical research. Existing optical chemical structure recognition (OCSR) tools fail to autonomously associate molecular structures with their bioactivity profiles, creating a critical bottleneck in structure-activity relationship (SAR) analysis. Here, we present BioChemInsight, an open-source pipeline that integrates: (1) DECIMER Segmentation and MolVec for chemical structure recognition, (2) Qwen2.5-VL-32B for compound identifier association, and (3) PaddleOCR with Gemini-2.0-flash for bioactivity extraction and unit normalization. We evaluated the performance of BioChemInsight on 25 patents and 17 articles. BioChemInsight achieved 95% accuracy for tabular patent data (structure/identifier recognition), with lower accuracy in non-tabular patents (~80% structures, ~75% identifiers), plus 92.2 % bioactivity extraction accuracy. For articles, it attained>99% identifiers and 78-80% structure accuracy in non-tabular formats, plus 97.4% bioactivity extraction accuracy. The system generates ready-to-use SAR datasets, reducing data preprocessing time from weeks to hours while enabling applications in high-throughput screening and ML-driven drug design (https://github.com/dahuilangda/BioChemInsight).

Problem

• Automated extraction of chemical structures and bioactivity data
• Linking molecular structures with bioactivity profiles autonomously
• Generating ready-to-use SAR datasets efficiently

Innovation

• Integrates DECIMER and MolVec for structure recognition
• Uses Qwen2.5-VL-32B for compound identifier association
• Combines PaddleOCR and Gemini-2.0-flash for bioactivity extraction