Dataset for the artificial neural network-based evaluation of ionization in supercritical fluid chromatography-mass spectrometry using electrospray (ESI) and atmospheric pressure chemical ionization (APCI)

Raw data used for the evaluation of ionization in supercritical fluid chromatography-mass spectrometry using electrospray (ESI) and atmospheric pressure chemical ionization (APCI) pubhlished in article "Artificial Neural Networks: An Innovative Approach Used for Elucidation of Ionization Processes in Supercritical Fluid Chromatography-Mass Spectrometry" in Analytical Chemistry, 2025. Data set contains: (i) 226 molecular descriptors calculated by CDK Descriptor Calculator (v.1.4.8) from 3D structures of the 107 analytes optimized by semi-empirical AM1 quantum mechanical calculations using the MOPAC application of Chem 3D Pro version 14.0 software (CambridgeSoft) (Excel sheet), (ii) MS responses (peak areas) for each compouned analyzed using SFC-ESI-MS with selected ion monitoring (SIM) of protonated and deprotonated molecular ions and using 224 make-up solvent compositions correlated to QC samples and splitting ratio, (iii) MS responses (peak areas) for each compouned analyzed using SFC-APCI-MS with selected ion monitoring (SIM) of protonated and deprotonated molecular ions and using 224 make-up solvent compositions correlated to QC samples and splitting ratio, (iv) weights assigned to each molecular descriptor based on each experimental conditions by artificial neural network created using the neural network simulator in Matlab R2023a with the deep learning toolbox V.23.2 (The MathWorks, Inc., Massachusetts, USA) and a sigmoid activation function, a backpropagation learning algorithm with 500 learning cycles (Excel sheet).本数据集用于评估采用电喷雾电离（electrospray, ESI）与大气压化学电离（atmospheric pressure chemical ionization, APCI）的超临界流体色谱-质谱联用技术中的电离过程，相关研究发表于2025年《分析化学（Analytical Chemistry）》期刊的论文《人工神经网络：用于解析超临界流体色谱-质谱联用技术电离过程的创新方法》。 该数据集包含以下内容：(i) 由CDK Descriptor Calculator（版本1.4.8）基于107种分析物的三维结构计算得到的226个分子描述符；上述分析物的三维结构通过Chem 3D Pro 14.0软件（CambridgeSoft公司）的MOPAC模块，采用半经验AM1量子力学计算完成优化，相关数据存储于Excel表格中；(ii) 采用超临界流体色谱-电喷雾质谱联用（SFC-ESI-MS）技术，对质子化与去质子化分子离子进行选择离子监测（selected ion monitoring, SIM）分析时，每种化合物的质谱响应值（峰面积），同时包含224种与质控（QC）样品及分流比相关的补充溶剂配比信息；(iii) 采用超临界流体色谱-大气压化学电离质谱联用（SFC-APCI-MS）技术，对质子化与去质子化分子离子进行选择离子监测分析时，每种化合物的质谱响应值（峰面积），同样包含224种与质控样品及分流比相关的补充溶剂配比信息；(iv) 基于各实验条件，通过人工神经网络为每个分子描述符赋予的权重；该人工神经网络依托Matlab R2023a软件及其深度学习工具箱V.23.2（美国马萨诸塞州The MathWorks公司）中的神经网络模拟器构建，采用sigmoid激活函数与反向传播学习算法，共执行500轮学习迭代，相关数据存储于Excel表格中。