It is very important to rapidly discover and identify the multiple components of traditional Chinese medicine (TCM) formula. High performance liquid chromatography with high resolution tandem mass spectrometry (HPLC-HRMS/MS) has been widely used to analyze TCM formula and contains multiple-dimension data including retention time (RT), high resolution mass (HRMS), multiple-stage mass spectrometric (MSⁿ), and isotope intensity distribution (IID) data. So it is very necessary to exploit a useful strategy to utilize multiple-dimension data to rapidly probe structural information and identify chemical compounds. In this study, a new strategy to initiatively use the multiple-dimension LC–MS data has been developed to discover and identify unknown compounds of TCM in many styles. The strategy guarantees the fast discovery of candidate structural information and provides efficient structure clues for identification. The strategy contains four steps in sequence: (1) to discover potential compounds and obtain sub-structure information by the mass spectral tree similarity filter (MTSF) technique, based on HRMS and MSⁿ data; (2) to classify potential compounds into known chemical classes by discriminant analysis (DA) on the basis of RT and HRMS data; (3) to hit the candidate structural information of compounds by intersection sub-structure between MTSF and DA (M,D-INSS); (4) to annotate and confirm candidate structures by IID data. This strategy allowed for the high exclusion efficiency (greater than 41%) of irrelevant ions in er-xian decoction (EXD) while providing accurate structural information of 553 potential compounds and identifying 66 candidates, therefore accelerating and simplifying the discovery and identification of unknown compounds in TCM formula.

快速发现和识别中药（TCM）配方的多个组成部分非常重要。高效液相色谱和高分辨率串联质谱（HPLC-HRMS / MS）已被广泛用于分析中药配方并包含多维数据，包括保留时间（RT），高分辨率质量（HRMS），多级质谱（MS ⁿ）和同位素强度分布（IID）数据。因此，非常有必要利用一种有用的策略来利用多维数据来快速探测结构信息并识别化合物。在这项研究中，已开发出一种主动使用多维LC-MS数据的新策略，以发现和鉴定许多样式的中药未知化合物。该策略可确保快速发现候选结构信息，并为识别提供有效的结构线索。该策略依次包括四个步骤：（1）基于HRMS和MS ⁿ，通过质谱树相似性过滤器（MTSF）技术发现潜在的化合物并获得子结构信息。数据; （2）根据RT和HRMS数据，通过判别分析（DA）将潜在化合物分类为已知化学类别；（3）通过MTSF与DA之间的交点子结构（M，D-INSS）命中化合物的候选结构信息；（4）通过IID数据对候选结构进行注释和确认。该策略可实现二仙汤（EXD）中不相关离子的高排除效率（大于41％），同时提供553种潜在化合物的准确结构信息并鉴定出66种候选化合物，从而加快并简化了未知化合物的发现和鉴定在中医配方中。