Abstract: As the definitions of 36 atomic fragment types in organic compounds，multi-order atom-pair frequency matrix was constructed in terms of atomic fragments occurring in pair at different bond distances，and based on a new molecular coding technique as characteristic atom-pair hologram code（CAHC）proposed in this paper. Collected from reference reports，a large-scale ion mobility spectrometry collision cross section database comprising 819 samples was established and quantitative structure-spectrometry relationship（QSSR）studies were performed with the CAHC. Testing modeling stabilities and generalization abilities by both internal and external examinations confirmed that CAHC was in obvious linear relationship with peptide collision cross sections，while it was involved in partially nonlinear factors for a few polypeptides. The model was deemed to assist in quantitative computer-aided predictions for peptide collision cross sections.

摘要： 通过定义有机化合物的36种原子碎片类型，并按分子中不同键距原子碎片成对出现次数构造多阶原子对频数矩阵，在此基础上得到了一种新的分子编码技术：特征原子对全息码(characteristic atom-pair hologram code, CAHC)。从文献报道收集组建了一个包含819个样本的大尺度肽离子迁移谱碰撞截面数据库，以此为基础采用CAHC成功实现了该组数据的定量构谱关系(QSSR)建模研究，通过内部及外部双重验证的办法对所得模型稳定性能及泛化能力进行了深入分析和检验，其结果表明CAHC码与肽离子碰撞截面呈显著线性相关，而对少数多肽则包含一定非线性因素。本文模型有助于实现计算机辅助肽离子碰撞截面定量预测自动化平台的构建。