Analysis of Differences in Different Parts of Cotton Flowers Based on Chemical Pattern Recognition and Entropy Weight TOPSIS Method
Cotton flowers are the dry flowers of the cotton plant Gossampinus malabarica (DC.) Merr. They are cool in nature and belong to the large intestine meridian. They have the effects of clearing heat and dampness, detoxifying, and are commonly used for diseases such as diarrhea, dysentery, hemorrhoids, and bleeding. Cotton flowers are abundant in resources and widely distributed in Yunnan, Hainan, Guangxi, Guangdong, Taiwan and other regions of China. They are usually harvested during the spring when the flowers are in full bloom. Modern research has shown that cotton flowers are rich in flavonoids, phenylpropanoids, phenolic acids, and other compounds. Among them, flavonoid compounds such as mangiferin and rutin have good therapeutic effects in the treatment of liver injury, myocardial ischemia-reperfusion injury, lung cancer, and other diseases; Phenolic acids have strong DPPH free radical scavenging activity, and protocatechuic acid has potential hypoglycemic effect in diabetes. At present, there are many studies on the microscopic identification, thin-layer identification, and mangiferin content determination of cotton flowers, but there are few reports on the HPLC fingerprint of cotton flowers and the quality differences of different parts.

There are certain limitations in evaluating the quality of kapok flower medicinal materials solely based on a single chemical component. The organic combination of traditional Chinese medicine fingerprint and chemometrics can comprehensively reflect the diversity and complexity of kapok flower chemical components, and systematically characterize the quality differences of different parts of kapok flower. The entropy weighted optimal solution distance (TOPSIS) method obtains a comprehensive index by reasonably weighting multiple indicators, and uses the degree of distance from the idealized target as the benchmark for comprehensive evaluation. It can effectively avoid the influence of subjective factors on the internal quality evaluation of cotton flowers, and has a certain degree of accuracy and scientificity. This study established HPLC fingerprint spectra of different parts (petals, sepals, stamens) of kapok flower medicinal materials. Similarity evaluation, heatmap clustering analysis, principal component analysis, and orthogonal partial least squares discriminant analysis were used to evaluate the quality of fingerprint spectra of different parts of kapok flower. The entropy weight TOPSIS method was used to comprehensively score and rank the quality of different parts of kapok flower. The aim is to provide data support for the enrichment of multiple indicator components in kapok flower medicinal materials and resource development, laying the foundation for its quality control and clinical application research.

 

When examining the chromatographic conditions, this study found that when the detection wavelength was 300nm, some flavonoid compounds did not exhibit corresponding UV absorption in their chromatographic peaks; When the detection wavelength is 255nm, there is a significant difference in the peak response values of each color spectrum. The chromatographic peak response of mangiferin is higher, and some chromatographic peaks have smaller absorption or loss; When the detection wavelength is 230nm, the response values of each spectral peak signal are relatively high, and the baseline is relatively stable. Therefore, 230nm is used as the detection wavelength. At the same time, different extraction methods (ultrasound, reflux), extraction solvents (methanol, 50% methanol, 70% methanol), solvent dosage (15, 25, 50mL), and extraction time (30, 45, 60min) were also investigated. The results showed that the sample solution prepared using the method under “2.3” had better peak separation and higher component content.
The research results indicate that the similarity between cotton flowers and sepals is relatively high, while the similarity between petals and stamens is relatively close. The results of heat map clustering analysis indicate that the quality consistency between petals and stamens is higher, and the chemical components of cotton flowers are mainly enriched in the sepals, which is consistent with the similarity results. The results of CA and PCA show that both can clearly characterize significant differences between sepals, petals, and stamen samples, while the differences between petals and stamens are relatively small. Combined with OPLS-DA analysis, it is possible to completely distinguish the petals, sepals, and stamens of cotton flowers. Peak 8, mangiferin, peak 13, peak 6, chlorogenic acid, peak 5, and rutin are the differential compounds between the three, which can be used as indicator components to distinguish different parts of cotton flower medicinal materials. At the same time, combined with the entropy weight TOPSIS method, the characteristic peaks of different parts of 15 batches of kapok flower medicinal materials were weighted, and the samples of different parts of kapok flower medicinal materials were sorted according to the Ci value, which can achieve overall control of kapok flower medicinal materials and screening of high-quality seed sources. The research results show that the average Ci value of cotton flower samples in Yunnan Province is higher than that of samples from other production areas, and the quality of sepals is better.
In summary, the HPLC fingerprint detection method established in this study for kapok flower medicinal materials and their different parts is stable and reliable. Through chemical pattern recognition and entropy weight TOPSIS method, the HPLC fingerprint of kapok flower medicinal materials and their different parts can be analyzed and evaluated comprehensively, comprehensively, and systematically. This allows for quality evaluation and differential analysis of samples, comparison of chemical composition differences in different parts, clarification of the distribution pattern of chemical components, and data support for the quality control and clinical application of kapok flower medicinal materials.