Comprehensive Analysis of Sinomenii caulis and Menispermi rhizoma Based on Ultra‐High Performance Liquid Chromatography and Ultra‐High Performance Liquid Chromatography‐Quadrupole Time‐of‐Flight‐Mass Spectrometry Authors: Ye Tao, Xinyue Zhang, Wen Li, Zhe Wu, Shuchen Duan, Lizhi Wan, Xiangri Li, Yazhong Zhang, Xianlong Cheng, Feng Wei Published: 2025 ABSTRACT ABSTRACTSinomenii caulis (SC) and Menispermi rhizoma (MR) are Chinese herbal medicines from the same family and are often confused. To achieve identification of SC and MR, the paper used the differential chemical components and the combination characteristics of chemical components to carry out a comprehensive analysis. First, the fingerprints of SC and MR were established separately using the ultra‐high performance liquid chromatography (UPLC) to screen the differential characteristic components. Furthermore, UPLC‐Quadrupole time‐of‐flight‐mass spectrometry (UPLC‐QTOF‐MS) analysis was carried out on SC and MR to acquire the digital matrix of chemical components (DMCCs) for characterizing SC and MR, respectively. Finally, the DMCCs of SC and MR were used for identification. The results of UPLC fingerprints showed good similarity among different batches of the same medicinal material. SC contains sinomenine (C19H23NO4), and MR contains dauricine (C38H44N2O6). A new identification method was further established based on UPLC‐QTOF‐MS and DMCC. This method successfully achieved the differentiation of SC from MR, and the detection results showed a correlation trend for samples with different adulteration ratios. Finally, with the MS (28%) of the sample adulterated with 3% MR as the limit, effective identification of 10 batches of blind samples was realized. Based on UPLC fingerprint and DMCC, the identification of SC and MR can be achieved with high quality and efficiency. It helps the authenticity identification of SC. Abbreviations digital matrix of chemical component Menispermi rhizoma match score Sinomenii caulis Introduction Sinomenii caulis (SC), known as “Qing Feng Teng”, is the dried vine stem of Sinomenium acutum (Thunb.) Rehd. et Wils. or Sinomenium acutum (Thunb.) Rehd. et Wils. var. cinereum Rehd. et Wils. [1, 2]. SC has a wide distribution range, found in Hubei, Anhui, Hunan, Zhejiang, Sichuan, Guangxi, Guizhou, and other regions in China [3]. In practical application, there are differences in medication practices and names between different regions, and there were wide, confusing varieties in circulation. Through actual market research, it was found that Menispermi rhizoma (MR), as one of SC's related species, was often confused with SC, and even SC was adulterated with MR in herbal markets [4]. However, the efficacy of SC and MR is different. SC can fight inflammation, dispel wind, remove dampness, and unblock meridians, while MR can fight inflammation, clear heat, and detoxicate [5, 6, 7]. Moreover, MR has certain toxicity, which poses a potential threat to personal safety. With the increasing medicinal and economic value of SC, the phenomenon of SC adulterated with MR in the medicinal market has become more and more frequent. Therefore, there is an urgent need to strengthen the regulation of SC and MR. In order to identify SC and MR, researchers have also carried out many exploratory analyses. Li et al. conducted morphological identification of SC and its adulterants, and the results showed that SC is an elongated cylindrical shape, often slightly curved, with a diameter of 0.5–2.0 cm, and greenish‐brown to brownish‐brown surface, which are very similar to SC's adulterants [8]. Shi and Song et al. carried out microscopic identification of SC and its adulterants, but the results showed they were highly similar in terms of their cork layer, vascular bundles, ray cells, stone cells, and pith [9, 10]. Wang et al. explored the quality markers of SC based on spectral efficacy relationships and chemometric analysis, and the results indicated that sinomenine, magnoflorine could be used for SC's quality control [11]. In addition, the Chinese Pharmacopoeia clearly stipulated that dauricine and daurisoline were the components used for identification and quality control [2, 11]. In summary, existing studies mainly focus on traditional methods (morphological and microscopic characteristics) and single markers to achieve the differentiation of SC and MR. Traditional identification methods rely heavily on empirical approaches, are inherently subjective, and often have ill‐defined discriminatory characteristics. Single‐marker identification focuses on a single chemical component; although it can objectively differentiate the two, it fails to explain the complexity and integrity of the medicinal material components and lacks effective analysis of adulterated samples. Ultra‐high performance liquid chromatography (UPLC) can rapidly separate components in complex samples, but its fingerprint chromatogram focuses on known compounds, easily overlooking unknown components with low response and thus limiting the comprehensive quality evaluation of natural products. UPLC‐mass spectrometry (UPLC‐MS) integrates high‐efficiency separation with high‐sensitivity and high‐resolution detection functions, and has become the core technology in this field. This technology enables accurate quantification of multiple bioactive components in natural products, such as the comparative determination of major components in Dioscorea nipponica and D. panthaica [12], the characterization and quantification of six flavonoids in “Dragon's Blood” [13], and the specific detection of diosgenin in Dioscorea species [14]. It can also be used for the identification of in vivo metabolites of natural products to clarify the in vivo transformation pathways of key substances [15]; combined with chemometric methods, it can distinguish closely related species [16, 17]. Its application in multi‐dimensional quality evaluation provides reliable support for the variety identification, quality standardization, and clinical application of natural products, promoting the development of the field toward precision. Focusing on the research gap regarding adulterated samples of SC, this study proposes a digital identification strategy based on UPLC and UPLC‐Quadrupole time‐of‐flight‐MS (UPLC‐QTOF‐MS): UPLC is employed to establish the fingerprints of SC and MR, with component identification as the core to achieve their differentiation; meanwhile, via UPLC‐QTOF‐MS, the digital matrices of chemical components (DMCCs) of the two are obtained, and a comprehensive digital identification method for SC and MR is established from the perspective of multi‐component differences. Consumables and Methods Consumables, Reagents, and Instruments UPLC Liquid Phase System (SHIMADZU Company); G2 Q‐TOF Mass Spectrometer (Waters, USA); Ultrasonic Cleaner (Kunshan Ultrasonic Instrument Co., Ltd.); XS‐105DU 100000 parts per million Balance (METTLER TOLEDO Company); Milli‐Q Pure Water Purifier (Merck Millipore Co., Ltd.); Sodium dihydrogen phosphate (Lot: 20180607, Sinopharm Chemical Reagent Co., Ltd.); Sodium formate (Lot: S30243, purity: 99.5%, Shanghai Yuanye Biotechnology Co., Ltd.); Leucine enkephalin (Lot:186006013, purity: 99%, Waters Technology Co., Ltd.); Sinomenine (Lot:110774‐202409, purity: 99.8%), magnoflorine (Lot: 112090‐202201, purity: 99%), and dauricine (Lot: 111867‐202205, purity: 97.2%) were supplied by NIFDC institution. 48 SC and 32 MR were bought from different origins by NIFDC. All samples were identified by Wei Feng as complying with pharmacopoeia requirements. Eight homemade samples with different percentages of adulteration (0%, 1%, 3%, 5%, 10%, 20%, 40%, 60%, 80%, and 100%, percentage of MR) [18]. In addition, 10 SC blind samples were collected from the herbal market for testing. Table S1 presents specific information about the sample materials. Sample Treatment Treatment of Chemical Reference Standards Weigh 20.00 mg of sinomenine, magnoflorine, and dauricine, respectively, and add 200 mL of methanol to make the concentration of 0.1 mg/mL of the chemical reference solution. Test Sample Treatment Sample powder of each batch of herbs was accurately weighed, 0.50 g (passed through sieve No. 4). Then, 30 mL of methanol was added for ultrasonic treatment for 30 min to obtain analysis samples [19]. Analysis Conditions UPLC Analytical Conditions The column was packed with octadecylsilane‐bonded silica gel (100 mm × 2.1 mm × 1.8 µm). Methanol was used as the A phase, and a 0.05 mol/L Na2HPO4 solution was used as the B phase. Table 1 shows the gradient elution. The temperature was 40°C. The analytical wavelength was set to 280 nm. TABLE 1The gradient elution procedure of ultra‐high performance liquid chromatography (UPLC) analysis.Time (minutes)Mobile phase A (%)Mobile phase B (%)Flow rate0–25→1095→90 0.2 mL/min 2–310→1990→813–619→2081→806–1020→2180→7910–2021→3079→7020–3030→4070→6030–30.140→560→9530.1–35595John Wiley & Sons, Ltd. The gradient elution procedure of ultra‐high performance liquid chromatography (UPLC) analysis. 0.2 mL/min MS Setting Parameters Using the Waters G2 QTOF instrument to perform MS analysis. Using Waters Acquity UPLC and BEH‐C18 to conduct chromatographic separations. Chromatographic temperature and analytical volume were 35°C and 2 µL. Table 2 shows mobile phase conditions. An electrospray ion source (ESI+), MSE acquisition method, whose scan rate was 0.2 s. m/z: 100 ∼ 1200. At the same time, 3.0 kV of capillary as well as 40 V of sampling cone. 450°C of desolvation temperatures with 800 L/h of desolvation gas; 0–40 V of collision energy and 120°C of source temperatures [14, 15, 16]. TABLE 2The mobile phase condition of ultra‐high performance liquid chromatography‐quadrupole time‐of‐flight‐mass spectrometry (UPLC‐QTOF‐MS) analysis.Time (min)Flow rateMethanol0.05 mol/L sodium dihydrogen phosphate0–2 0.2 mL/min 5→1095→902–310→1990→813–619→2081→806–1020→2180→7910–2021→3079→7020–3030→4070→6030–30.140→560→9530.1–35595John Wiley & Sons, Ltd. The mobile phase condition of ultra‐high performance liquid chromatography‐quadrupole time‐of‐flight‐mass spectrometry (UPLC‐QTOF‐MS) analysis. 0.2 mL/min Data Processing Methods The UPLC chromatograms of SC and MR were processed using a similarity evaluation system. The MS of SC and MR were imported into QI software, sorted according to m/z, tR, maximum abundance (I), corrected for peaks, extracted, and then exported to quantified data for storage. Then, as shown in Figure 1, the common ions were isolated from multiple batches of SC and MR, respectively. Furthermore, the SC and MR's common ions were used for eliminating intersecting ion data to extract the top‐100 ions as SC and MR “DMCC”, respectively. In the aforementioned process, if ΔtR ≤ 0.10 min and Δm/z ≤ 0.03 Da, these ions are extracted or eliminated as common ions. The “DMCCs” of SC and MR were employed as a reference for identification, yielding a match score (MS) as feedback (MS = ion quantity matched successfully/100 × 100%). Finally, chemometric analysis was conducted in SIMCA 14.1 and used to analyze and verify the identification results based on “DMCC” matching [20]. The acquisition process for the digital matrix of chemical components of Sinomenii caulis and Menispermi rhizoma. Results and Discussions UPLC Analysis Both SC and MR were subjected to the same sample pretreatment and LC conditions, and the corresponding chromatograms were obtained. The overlaid chromatograms are shown in Figures 2 and 3. As indicated in Figures 2 and 3, the batches of the same medicinal material exhibited good similarity, confirming the feasibility of extracting “DMCC”. SC contains sinomenine (C19H23NO4), and MR contains dauricine (C38H44N2O6), which can serve as identification markers for the two. The UPLC method can only characterize the two‐dimensional information (tR; peak area) of the chemical composition, and a single marker provides limited information, resulting in significant limitations in identifying adulteration solely relying on it. Therefore, a combined method of MS and “DMCC” was further adopted to extract the three‐dimensional fingerprint information (tR; I; m/z) of SC and MR, and identification was performed based on multi‐component differences. The ultra‐high performance liquid chromatography (UPLC) schematic diagram of Sinomenii caulis fingerprint pattern. The ultra‐high performance liquid chromatography (UPLC) schematic diagram of Menispermi rhizoma fingerprint pattern. During the development of the UPLC analysis method, various parameters were optimized, including: sample extraction method (ultrasound, heating reflux, microwave extraction, steaming), extraction solvent (methanol, water, 80% methanol, acetonitrile), extraction time, and LC conditions (0.01 mol/L NaH2PO4, 0.02 mol/L NaH2PO4, and 0.05 mol/L NaH2PO4). The final optimized protocol was determined as follows: methanol, as a solvent with ultrasonic extraction for 30 min, proved most suitable. Methanol‐0.05 mol/L NaH2PO4 solution with gradient elution. Column temperature: 40°C. Detection wavelength: 280 nm. UPLC‐QTOF‐MS Analysis As shown in Figure 4, the MS of SC and MR exhibits certain similarities, and the blank did not interfere with detection. Simply comparing MS spectra makes it difficult to achieve identification between SC and MR, let alone adulteration analysis. Therefore, the “DMCC” of SC and MR was further proposed to solve the adulteration question and realize digital analysis. The base peak ion chromatogram of blank, Sinomenii caulis (SC), Menispermi rhizoma (MR), positive adulterated sample (1% MR, 3%MR), and blind sample. Before the MS detection of SC and MR, we optimized the treatment method and analytical conditions. In pure water, methanol, ethanol, acetonitrile, and 70% methanol‐water, it was indicated that more MS information on chemical composition was obtained when methanol was the solvent [21]. The processing time was examined for 10, 30, and 40 min, and finally, ultrasonic treatment with methanol for 30 min was more appropriate, saving time and more MS information. In MS analysis, different ion modes (positive and negative), and different collision energies (10–30, 10–40, and 10–50 eV) were examined, and it indicated that there was more MS data in the positive ion mode with collision energies of 10–40 eV [22]. Data Quantization and Digital Matrix of Chemical Components Five methanol samples were taken as “blanks” to deduct the interference ions, which contain 1588∼1658 [tR‐m/z‐I] units, with a mean value of 1623 and an RSD of 3.05%. Table S2 shows the quantity of [tR‐m/z‐I] of SC and MR used to construct “DMCC”. Table S2 shows that there are differences in the quantity of [tR‐m/z‐I] obtained between SC and MR, which indicates that the influencing factors of different origins, different ages, and individual differences led to the large differences in chemical compositions between batches, and taking such samples into consideration comprehensively yields more representative results and holds practical significance for testing. According to 2.4 data processing methods, all batches of SC and MR in Table S2 were calculated to obtain “DMCCs” of SC and MR. Tables S3 and S4 show the “DMCCs” of SC and MR, respectively. As shown in Tables S3 and S4, essentially, the “DMCCs” of SC and MR were the quantitative combination characteristics of the proprietary ions of SC and MR, respectively. At the same time, tR and m/z deviations were determined based on actual detection conditions. The maximum tR drift of the samples was 0.1 min, and the m/z deviation is less than 0.03 Da. Therefore, the threshold was ultimately determined to be ΔtR≤0.10 min and Δm/z≤0.03 Da. Identification of Sinomenii caulis and Menispermi rhizoma Table 3 shows the MS of SC test samples compared with the “DMCC” of SC. The MSs of most SC test samples were greater than 80% except the SC01, SC43, and SC45. The MSs of SC12 and SC22 were all 101%, which was caused by data bias and in line with actual circumstances. Table 4 shows the MS results of SC test samples compared with the “DMCC” of MR. Compared with MR's “DMCC”, the MSs of all SC test samples were all less than 10%. As shown in Tables 3 and 4, the smallest MS of SC45 compared with SC's “DMCC” was 61% and the maximum MS of SC45 compared with MR's “DMCC” was 8%, which had a difference of more than six times. In addition, Table 5 presents the nonparametric statistical analysis of the MS results from Tables 3 (Group X) and 4 (Group Y). p < 0.01 indicates that the MSs of SC compared with SC's “DMCC” differ markedly from the MSvalues of SC compared with MR's “DMCC”. TABLE 3The trusted values of Sinomenii caulis compared with the “DMCC” of SC.Herb nameBatchesMatching successful ionsIon number in “DMCC” of SCTVSinomenii caulis (SC)SC0170100‐SC70%SC0392100‐SC92%SC0686100‐SC86%SC0892100‐SC92%SC1085100‐SC85%SC12101100‐SC101%SC1593100‐SC93%SC1790100‐SC90%SC1997100‐SC97%SC22101100‐SC101%SC2495100‐SC95%SC2589100‐SC89%SC2791100‐SC91%SC2892100‐SC92%SC3094100‐SC94%SC3190100‐SC90%SC3391100‐SC91%SC3499100‐SC99%SC3789100‐SC89%SC3995100‐SC95%SC4187100‐SC87%SC4372100‐SC72%SC4561100‐SC61%John Wiley & Sons, Ltd. The trusted values of Sinomenii caulis compared with the “DMCC” of SC. TABLE 4The trusted values of Sinomenii caulis compared with the “DMCC” of SC.Herb nameBatchesMatching successful ionsIon number in “DMCC” of MRTVSinomenii caulis (SC)SC018100‐MR8%SC031100‐MR1%SC060100‐MR0%SC086100‐MR6%SC105100‐MR5%SC121100‐MR1%SC151100‐MR1%SC171100‐MR1%SC192100‐MR2%SC221100‐MR1%SC241100‐MR1%SC250100‐MR0%SC270100‐MR0%SC280100‐MR0%SC300100‐MR0%SC312100‐MR2%SC330100‐MR0%SC340100‐MR0%SC371100‐MR1%SC390100‐MR0%SC410100‐MR0%SC430100‐MR0%SC450100‐MR0%John Wiley & Sons, Ltd. The trusted values of Sinomenii caulis compared with the “DMCC” of SC. TABLE 5The nonparametric statistical analysis for trusted values of Sinomenii caulis (SC).ClassMedianMann Whitney‐UMann Whitney‐Z p Group X (n = 23)Group Y (n = 23)TV91%1%0−5.8630.000 ** * p < 0.05.** p < 0.01.John Wiley & Sons, Ltd. The nonparametric statistical analysis for trusted values of Sinomenii caulis (SC). p < 0.05. p < 0.01. On the other hand, Figure 5 shows the MS of SC test samples compared with the “DMCCs” of SC and MR. The MSs of all MR test samples compared with MR's “DMCC” were greater than 67% while the MSs of all MR test samples compared with SC's “DMCC” were less than 3%, which had a difference of more than twenty times. Table 6 shows the nonparametric statistical analysis of the MS results of MR (Group M: the MSs of MR compared with its own “DMCC”; Group N: the MSs of MR compared with SC's “DMCC”), which indicates that there is a huge difference in TVs. To sum up, the “DMCCs” of SC and MR have certain specificity, and the MSs obtained by the matching of SC and MR's “DMCCs” can effectively and accurately identify SC and MR. The matching situation of Menispermi rhizoma. TABLE 6The nonparametric statistical analysis for trusted values of Menispermi rhizoma (MR).ClassMedianMann Whitney‐UMann Whitney‐Z p Group M (n = 17)Group N (n = 17)TV79%1%0−5.0200.000*** p < 0.05.** p < 0.01.John Wiley & Sons, Ltd. The nonparametric statistical analysis for trusted values of Menispermi rhizoma (MR). p < 0.05. p < 0.01. Identification of Adulterated Samples Figure 6 shows the matching results of adulterated samples with “DMCCs” of SC and MR. As the proportion of MR in the adulterated samples increased, the MS between adulterated samples and MR's “DMCC” exhibited an upward trend. When compared with MR's “DMCC”, from 20% to 100% MR, the MSs of adulterated samples were all greater than 65%. Even if 1% MR, the MS reached 17% when matched with the MR's “DMCC”. However, the 0% MR showed an MS of 0%. On the other hand, as the proportion of SC in the adulterated samples decreased, the MS between adulterated samples and SC's “DMCC” also exhibited a downward trend. When compared with SC's “DMCC”, from 0% to 60% MR, the MSs of adulterated samples were not less than 68%. At the same time, the MS between 80%MR and SC's “DMCC” was only 38% and the MS between 100%MR and SC's “DMCC” was only 1%, which is a difference of more than 30 times. On the whole, the “DMCCs” of SC and MR have a certain proprietary and can identify adulterated samples by comparing the MSs. The matching situation of adulterated samples. 0212 general rules of Ch.P four parts stipulate that the limit of impurities in TCM is generally 3%. Therefore, the MS (28%) value of 3%MR adulterated samples can be selected as the impurity detection limit to further identify SC blind samples. Analysis of SC Blind Samples Figure 7 shows the matching results between the SC blind samples and the “DMCC” of MR. As shown in Figure 7, among the 10 batches of blind samples, except for AS06, the MSs between the remaining blind samples and “DMCC” of MR were all below 6%, which is far below the detection threshold of 28%. At the same time, the MS between AS06 and “DMCC” of MR was 65%, which is far greater than 28%. Therefore, it can be proven that AS06 was doped with MR, and other SC blind samples were not doped with MR. In addition, Figure 8 shows the matching results between the SC blind samples and the “DMCC” of SC. The MSs of all the SC blind samples were higher than 70%, which indicated that it does not affect the identification of SC. Further, the stoichiometry was used to obtain proprietary ions for identifying SC blind samples [23]. Figure S1 shows the results of chemometric analysis. As shown in Figure S1A, based on two principal components, MR was distinctly different from SC with Q2 = 0.552, which indicated there were different chemical components. The OPLS‐DA analysis in Figure S1B showed that MR can also significantly distinguish SC in supervised mode with Q2 = 0.989. Furthermore, the S‐plot analysis in Figure S1C showed the specificity ion in MR (red dot). Following comparison with chemical reference standards and literature review, the proprietary chemical composition of MR is dauricine (6.06 min_m/z 313.18, [M+2H]+). Therefore, on the other hand, the presence of MR in the blind sample can be determined by extracting the characteristic ions of MR. Figure 9 shows the ion extraction status. The proprietary ion (6.06 min_m/z 313.18, [M+2H]+) can be detected in the MR, 1%MR adulterated sample, AS06, and cannot be detected in SC and the remaining blind samples. It proves that the AS06 blind sample was indeed adulterated with MR. At the same time, it also demonstrates that the “DMCC”‐based identification method was correct. The matching results between the Sinomenii caulis blind samples and the “digital combination of chemical components” of Menispermi rhizoma. The matching results between the Sinomenii caulis blind samples and the “digital combination of chemical components” of Sinomenii caulis. The extraction situation for the proprietary chemical component of Menispermum rhizoma – dauricine. Features and Advantages The distinctive feature of this study lies in proposing the “DMCC” for SC and MR. Its essence is a quantitative combination characteristic of chemical composition. Compared to fingerprinting technology and single‐component identification, DMCC enables the effective and rational utilization of unknown chemical components without requiring the identification of specific compounds, which facilitates digital analysis in the era of digital and intelligent traditional Chinese medicine. As for the identification results, based on fingerprint patterns, it is difficult to identify adulterated samples due to the high similarity of fingerprint patterns. Although adulteration analysis based on characteristic chemical components is feasible, it incurs high analytical costs, requires lengthy testing times, and is susceptible to sample matrix effects that compromise the reproducibility of results. Furthermore, it is difficult to implement for rapid on‐site screening, poorly suited for high‐throughput testing of bulk samples, and thus fails to meet the practical needs of real‐time monitoring and rapid analysis. However, under the condition of ensuring the stability of data reproduction, the “DMCC” does not need to identify chemical components and can realize the detection and supervision of large quantities of samples quickly. Moreover, the “DMCC” (combined characteristics of multi‐component systems) exhibits greater specificity compared to those of single‐component systems. Application Exploration Based on the comparison of the UPLC chromatogram, the identification of MR and SC in a small batch can be realized. However, UPLC has insufficient ability to identify adulterated samples, especially when a small amount of MR is added. Furthermore, as for adulterated samples, the “DMCCs” of SC and MR were proposed and constructed based on MS analysis. The “DMCCs” of MR and SC are essentially an ordered combination of quantitative data about the ion information of chemical components. Compared to a single chemical component, it possesses greater specificity, and it is more suitable for the digital preliminary analysis of large‐scale samples, which can enhance analytical capabilities and efficiency [24, 25]. Due to data variability, when conducting adulteration analysis based on “DMCCs” of MR and SC, a 3% adulteration control sample should be prepared. For samples whose preliminary digital identification based on “DMCCs” of MR and SC is inconclusive, the precise matching analysis can be achieved through the extraction of specific ions. In summary, the “DMCCs” of MR and SC demonstrate strong practicality and facilitate adulteration analysis for MR and SC. It provides an effective way and method to realize digital regulation of MR and SC. Conclusion Based on UPLC and UPLC‐QTOF‐MS, a comprehensive analysis of SC and MR, especially the adulteration analysis, has been successfully realized. At the same time, the “DMCCs” of SC and MR have certain specificity, which helps to realize the digital analysis of SC and MR with MS≥ 60%. Even when 3% MR was mixed in SC, it could be detected with high quality and efficiency. In addition, one SC blind sample was adulterated with MR. It helps the authenticity identification of SC. Author Contributions Ye Tao: writing‐original draft, writing‐review & editing, validation, and conceptualization. Xinyue Zhang: writing‐original draft, writing‐review & editing, methodology, and investigation. Wen Li: Writing‐original draft, visualization, data curation, and software. Zhe Wu: software, data curation, investigation, and writing‐original draft. Shuchen Duan: investigation, supervision, validation, and data curation. Lizhi Wan: conceptualization, writing‐review & editing, validation, and investigation. Xiangri Li: supervision and project administration. Yazhong Zhang: funding acquisition, supervision, and project administration. Xianlong Cheng: writing‐review & editing and project administration. Feng Wei: funding acquisition, writing‐review & editing, project administration, and software. Funding The study was supported by State Key Laboratory of Drug Regulatory Science [2025SKLDRS0341] and the National Key Research & Development Program of China [2023YFC3504105] and Key Project of Regulatory Science Research of Anhui Drug Administration [AHYJ‐KJ‐202402]. Conflicts of Interest The authors declare no conflicts of interest. Supporting information Supporting File: ansa70058‐sup‐0001‐SuppMat.docx