A comprehensive characterization strategy for natural products using dual-mode hybrid acquisition on high-resolution MS with intelligent subdivision polygonal mass defect filtering: a case study on Typhae Pollen.

The polygonal mass defect filtering (PMDF) technique is widely used for initial filtering of characteristic constituents in natural products. This method still has issues with false positives in blank areas, a complex construction process, and cumbersome target components filtering. The intelligent subdivision polygonal mass defect filtering (Sub-PMDF) technique was developed to solve these issues. Python was introduced to automate both the construction of polygonal regions and the filtering of target components. The concept of subdivision was incorporated to reduce the blank areas within polygons, thereby lowering the rate of false positives. Furthermore, the chemical information of samples was obtained by a dual-mode hybrid acquisition combining progressive data-dependent acquisition with multi-window data-independent acquisition. MS-DIAL, GNPS, and CFM-ID were integrated to automatically characterize these compounds. The characteristic diagnostic ions, neutral loss fragments, and local databases matching were used for manually verifying the accuracy of the characterization of these compounds. A total of 130 compounds were identified in Typhae Pollen: 43 flavonoids, 26 fatty acids, 12 phenolic acids, 10 organic acids, 4 alkaloids, and 35 other compounds. It was concluded that dual-mode hybrid acquisition on high-resolution MS with intelligent Sub-PMDF is suitable for the chemical characterization of natural products and has higher accuracy and efficiency, broader coverage and fewer false positives.