Knowledge about the 3-dimensional structure, orientation and interaction of chemical compounds is important in many areas of science and technology. X-ray crystallography is one of the experimental techniques capable of providing a large amount of structural information for a given compound, and it is widely used for characterisation of organic and metal-organic molecules. The method provides precise 3D coordinates of atoms inside crystals, however, it does not directly deliver information about certain chemical characteristics such as bond orders, delocalization, charges, lone electron pairs or lone electrons. These aspects of a molecular model have to be derived from crystallographic data using refined information about interatomic distances and atom types as well as employing general chemical knowledge. This publication describes a curated automatic pipeline for the derivation of chemical attributes of molecules from crystallographic models. The method is applied to build a catalogue of chemical entities in an open-access crystallographic database, the Crystallography Open Database (COD). The catalogue of such chemical entities is provided openly as a derived database. The content of this catalogue and the problems arising in the fully automated pipeline are discussed, along with the possibilities to introduce manual data curation into the process.

中文翻译：

---

从晶体学数据导出化学实体的工作流程及其在晶体学开放数据库中的应用

关于化合物的三维结构、取向和相互作用的知识在许多科学技术领域都很重要。 X射线晶体学是能够为给定化合物提供大量结构信息的实验技术之一，广泛用于有机和金属有机分子的表征。该方法提供了晶体内部原子的精确 3D 坐标，但是，它不能直接提供有关某些化学特性的信息，例如键序、离域、电荷、孤电子对或孤电子。分子模型的这些方面必须使用有关原子间距离和原子类型的精确信息以及利用一般化学知识从晶体学数据中得出。该出版物描述了一个精心设计的自动管道，用于从晶体学模型推导分子的化学属性。该方法用于在开放访问晶体学数据库晶体学开放数据库（COD）中构建化学实体目录。此类化学实体的目录作为派生数据库公开提供。讨论了该目录的内容和全自动管道中出现的问题，以及在流程中引入手动数据管理的可能性。