SureChEMBL 2026 版发布：专利文献化学与生物医学注释开放数据库完成重构
SureChEMBL in 2026: an open database of chemical and biomedical annotations from the patent literature
SureChEMBL 发布 2026 版重大重构，在原有专利化学信息自动抽取的基础上新增基因、蛋白、疾病和机制等生物医学注释，把专利化学与生物语境关联起来。新版采用现代架构，重新设计了抽取、化学注册、检索和数据交付流程，并用统一标准化流程对历史语料做了大规模重处理。数据库目前收录 1.22 亿件专利、4400 万份带化学注释的文档、3100 万个化学结构和 100 万条生物医学实体注释，可通过网站、API 和 Bulk Data 下载获取。作者指出专利文献体量大、语言与格式异构、化学结构常不可机读，是规模化分析的主要障碍。
来源：Nucleic Acids Research · doi.org
数据库重构后新增基因、蛋白、疾病与机制注释，把专利化学信息与生物语境连起来，做药物发现数据挖掘的读者可据此判断是否迁移。
The patent literature is a major source of scientific and technical information for drug discovery, often disclosing chemical structures, biological targets, therapeutic indications, and structure-activity relationships before they appear in the scientific literature. However, patent documents are difficult to analyse at scale because of their volume, legal and technical language, heterogeneous formats, and frequent use of nonmachine-readable representations for chemical structures. Since 2015, SureChEMBL has provided open access to chemical information automatically extracted from patent documents. Here, we present a major redesign of the SureChEMBL resource. The updated platform is built on a modern architecture and includes redesigned extraction, chemistry registration, search and data-delivery workflows, together with large-scale reprocessing of the historical corpus through a unified standardization pipeline. SureChEMBL also introduces automatically generated biomedical annotations for genes, proteins, diseases, and mechanisms, linking patent chemistry to biological context. The resource currently contains 122 million patents, 44 million chemically annotated documents, 31 million chemical structures, and 1 million biomedical entity annotations, accessible through https://www.surechembl.org/, an API and the downloadable Bulk Data.