To establish a quality evaluation method for Chaihu Yujinxiang granules based on fingerprint, multi-component content determination, and chemical pattern recognition, providing important basis for its quality control.

The “Chinese Medicine Chromatographic Fingerprint Similarity Evaluation System (2012 Edition)” was used to establish HPLC fingerprints of 10 batches of Chaihu Yujinxiang granules, identify common peaks, and perform similarity evaluation. HPLC method for determining the content of four active ingredients, namely liquiritin apioside, liquiritin, liquiritigenin, and isoliquiritigenin was established. Cluster analysis and orthogonal partial least squares discriminant analysis were conducted on Chaihu Yujinxiang granules using SPSS 26.0 and SIMCA 14.1 software. Differential components affecting the quality of Chaihu Yujinxiang granules were screened based on the criterion of variable importance projection (VIP) value>1.0.

The established fingerprint and multi-component content determination method achieved satisfactory results after methodological investigation. The similarity of the fingerprint of 10 batches of Chaihu Yujinxiang granules ranges from 0.953 to 0.977, with a total of 11 common peaks, 4 of which were identified. The average contents of active ingredients such as liquiritin apioside, liquiritin, liquiritigenin, and isoliquiritigenin in 10 batches of Chaihu Yujinxiang granules were 2.71 mg·g^-1, 10.17 mg·g^-1, 2.47 mg·g^-1, and 1.86 mg·g^–1, respectively. The results of cluster analysis and principal component analysis indicates that 10 batches of Chaihu Yujinxiang granules could be clustered into two categories, with S3, S6, and S8 in one category and the rest in another category. The VIP values of peaks 2, 9, 7, and 3 were above 1.0.

The established fingerprint and content determination method are stable and reliable. Combined with chemical pattern recognition technology, they can be used to evaluate the overall quality of Chaihu Yujinxiang granules. Peaks 2, 9, 7, and 3 are differential components that affect the quality of Chaihu Yujinxiang granules.

To establish a liquid chromatography-tandem mass spectrometry (LC-MS/MS) method for determining the concentration of cyclic guanosine monophosphate (cGMP) in lysates of human colon adenocarcinoma lung metastasis cells (T84 cells) after co-incubation with linaclotide.

A Kromasil 100-5-C₁₈ column (150 mm×2.1 mm, 5 μm) was used with a mobile phase consisting of 0.1% formic acid in water and 0.1% formic acid in methanol, employing a gradient elution at a column temperature of 50 ℃. The detection was performed using an electrospray ionization (ESI^-) source and multiple reaction monitoring (MRM) mode, with the monitored ion transitions for cGMP and the internal standard 8-Br-cGMP being m/z 344.20 → 150.00 and m/z 423.90 → 230.00, respectively.

The linear ranges for cGMP were 1 to 500 ng·mL^-1 (r≥ 0.999). The method demonstrated precision, accuracy, matrix effects, and extraction recovery rates that met analytical requirements. The method was successfully applied to accurately detect cGMP levels in cells after administration of two types of linaclotide capsule formulations. A significant concentration-dependent change in cGMP levels was observed after co-incubation of linaclotide with T84 cells for 30 min, with EC₅₀ values of 167.6 and 112.1 nmol·L^-1 for the reference and test formulations, respectively.

The method established in this study demonstrates excellent selectivity and accuracy, effectively quantifying cGMP levels in lysates of human colon adenocarcinoma lung metastasis cells, providing a reliable analytical tool for related pharmacological research.

To establish a method combining size exclusion chromatography (SEC) with an optimized Infogest in vitro static digestion model for determining the relative molecular mass distribution of short-peptide-based enteral nutrition before and after digestion. To evaluate and assess the quality of three different products, to explore product quality characteristics and digestive stability, and to provide guidance for quality control and clinical individualized application of short-peptide-based enteral nutrition.

The SEC method was performed using two ECOSIL SEC G 2000 columns (300 mm×7.8 mm, 5 μm) in series, with acetonitrile-water-trifluoroacetic acid(15：8 5：0 .1) as the mobile phase, a flow rate of 0.7 mL·min^-1, the column temperature 30 ℃, the detection wavelength 215 nm, and the sample volume 20 μL. The in vitro simulated digestion process referred to and optimized the Infogest in vitro static digestion model, divided into simulated gastric digestion, simulated intestinal digestion, and simulated gastro-intestinal total digestion stages. The established SEC method was used to compare changes in the relative molecular mass distribution before and after digestion to evaluate in vitro digestive stability.

Within the range of relative molecular mass from 165.19 to 12 327, the logarithms of the relative molecular mass of the 8 reference standards showed good linear relationships with retention time. Specificity tests indicated that blank solvent and the simulated gastric-intestinal digestion solution had minimal impact on the measurement of relative molecular mass distribution. The RSDs for precision, repeatability, and 24 h stability tests were all less than 1.0%. The relative molecular mass of various short peptide enteral nutrition products was mainly concentrated between 150 and 1 000, but there were significant differences in the distribution ratios of relative molecular mass among different products. The hydrolysis of peptide components primarily occurred during the production stage, and after in vitro simulated digestion, their relative molecular mass distribution characteristics remained largely unchanged, with the degree of change mainly related to the original hydrolysis level of the product.

The SEC method combined with the optimized Infogest model is suitable for studying the molecular mass distribution and digestive stability of short-peptide-based enteral nutrition. The method is straightforward and rapid, facilitating a comprehensive assessment of the products quality characteristics and nutritional value.

To establish an LC-MS/MS method for determining the concentrations of imipenem and cilastatin in human plasma, for monitoring clinical therapeutic drug concentrations, and to investigate the effects of adding stabilizers during the sample pretreatment on mass spectrometry signal intensity.

After protein precipitation, the sample was subjected to gradient elution using an Agilent TC-C₁₈ (2) (150 mm×4.6 mm, 5 µm) column with a mobile phase system of 0.15% formic acid in water and methanol. The electrospray ionization (ESI) mass spectrometer was operated in positive ion mode using multiple reaction monitoring (MRM): m/z 300.1 → 141.9 (imipenem),m/z 359.7 → 97.0 (cilastatin) and m/z 384.1 → 141.1 (meropenem, internal standard). The samples containing and without 3-(N-morpholino) propane sulfonic acid (MOPS) as stabilizers were pretreated and continuously analyzed to compare the changes in mass spectrometry signal intensity.

Both imipenem and cilastatin showed good linearities in the concentration ranges of 0.1-100.0 μg·mL^-1 (r>0.99). The intra-day and inter-day accuracy ranges from 95.3% to 108.5%, the precision (RSDs) were less than 9.3%, the extraction recovery rate ranges from 77.4% to 84.3%, and the matrix effect ranges from 97.1% to 111.2%. Imipenem in plasma samples was stable at room temperature for 3 h, at 4 ℃ for 6 h, and at -80 ℃ for 12 d, while it was significantly degraded at -20 ℃ for 12 d. Cilastatin was stable under a variety of conditions. The method was robust to changing conditions of column temperature ±5 ℃, flow rate ±0.1 mL·min^-1, formic acid concentration in the aqueous phase ±0.025%, and ion source temperature ±50 ℃. The samples containing stabilizers exhibited significant ion inhibition on mass spectrometry after continuous injection, while samples without stabilizers had no significant effect on the signal intensity of mass spectrometry.

The method is simple and accurate and can be used for clinical drug monitoring of imipenem and cilastatin. Nonvolatile salt stabilizers such as MOPS can reduce mass spectrometry sensitivity, and the absence of such stabilizers is more suitable for long-term analysis by LC-MS/MS.

To systematically analyze the pharmacokinetic parameters, oral bioavailability and in vivo metabolites of trilobatin in Sprague-Dawley (SD) rats using liquid chromatography - triple quadrupole mass spectrometry (LC-MS/MS).

The chromatographic conditions were performed on an ACQUITY UPLC BEH C₁₈ column (50 mm×2.1 mm, 1.7 µm) with 0.1% formic acid (mobile phase A) and acetonitrile (mobile phase B) as the mobile phase. A gradient elution program was carried out with an accompanying flow rate of 0.3 mL·min^-1, a column temperature of 40 ℃, and an injection volume of 2 μL. The mass spectrometry conditions comprised an electrospray ion source in conjunction with negative ionization mode, with an ionogenic temperature of 150 ℃, a capillary voltage of -3.0 kV, and a desolvation-gas flow temperature of 500 ℃. The desolvation-gas flow rate was set at 750 L·h^-1, and the conical pore gas volumetric flow rate was fixed at 150 L·h^-1. The analysis was conducted in multiple reaction monitoring mode. Trilobatin was given to rats via gavage and intravenous injection, respectively. Plasma, urine and fecal samples were collected, and the drug concentration was determined after methanol precipitation of proteins. Pharmacokinetic parameters and metabolites were analyzed by pharmacokinetic software and metabolite analysis and identification software.

Following the administration of trilobatin to SD rats at a dose of 100 mg·kg^-1 via gavage and intravenous injection, respectively. The area under the curve (AUC_0-t) was found to be (423.98 ± 295.42) ng·h·mL^-1 and (90 894.75 ± 25 472.44) ng·h·mL^-1, respectively. The oral bioavailability was determined to be 0.46%; C_max was (203.83±25.88) ng·mL^-1 and (181 814.90±113 461.60) ng·mL^-1, respectively. The oral half-life was 1.65 h, while the intravenous half-life was 3.82 h. Trilobatin was metabolized to phloretin in the intestine and underwent further biotransformation in vivo through deglycosylation, methylation, deoxygenation and hydrolysis.

The pilot study represents a preliminary investigation into the in vivo pharmacokinetics and metabolism of trilobatin in rats, providing a foundation for further pharmacodynamics research and subsequent formulation development.

To establish a principal component external standard method with HPLC and calibration factors for the determination of 11 impurities in flurbiprofen axetil injection, and to explore its detection results and limit values.

The Thermo BDS Hypersil C₁₈ (250×4.6 mm, 5 μm) was selected for gradient elution, with water-0.15% acetic acid and acetonitrile-0.15% acetic acid as the mobile phase at a flow rate of 1.0 mL·min^-1. The column temperature was 40 ℃, the detection wavelength was 254 nm and the injection volume was 10 μL.

Flurbiprofen axetil and 11 impurities were well separated by the method. Good linearity was obtained with correlation coefficients of 1.000 for the 3-fluoro-4-phenylphenol (4-OHB), 1-acetoxyethyl-2-(2-fluoro-4-biphenylyl)-2-hydroxypropionate (2-OHP), 4-acetyl-2-fluorobiphenyl (4-ACB), flurbiprofen ethyl ester, allyl -(2-fluoro-4-biphenyl) propionate (ALE), ChP impurityⅠ , impurity B, impurity C and impurity E in the range of 0.10-20 μg·mL^-1. The average recovery rates was from 96.6% to 103.7% and the relative standard deviations(RSDs) were lower than 1.4%. The correction factors of flurbiprofen axetil related substances 4-OHB, 2-OHP,4-ACB, flurbiprofen, flurbiprofen ethyl ester, ALE, ChP impurity Ⅰ, impurity B, impurity C and impurity E were 0.55, 1.05, 1.01, 0.76, 0.95, 0.86, 0.55, 0.93, 0.76 and 0.81, respectively. Notablely, desfluoro fiurbiprofen axetil of detected was around the prescribed limit 0.1%.

The method above is rapid, simple, accurate, and reliable, and can be applied for the determination and quality control of related substances in flurbiprofen axetil injection.

To identify the authenticity of four batches of “Ophiopogonis Radix” in the market by means of multiple means, and to explore the reasons for exceeding the limit of its phloem bundles, so as to provide evidence for its inspection and detection.

Based on the relevant provisions of Ophiopogon japonicus (L. f.) Ker-Gawl., a variety in the 2020 edition of Chinese Pharmacopoeia, comparing the common confused products of Ophiopogon japonicus (L. f.) Ker-Gawl., Combined with traditional identification methods (character identification, microscopic identification) and modern analysis techniques (molecular biology ITS 2 sequence), the data of genuine products with mixed products and substandard phloem bundle samples were campared and analyzed.

The differences between Liriope spicata (Thunb.) Lour. and Ophiopogon japonicus (L. f.) Ker-Gawl. were the surface color, the depth of vertical wrinkles and the thickness of the middle column, and the differences in microscopic cross-sections lay in the number of phloem bundles and whether the inner cortex cells were uniformly thickened. The four batches of “Ophiopogonis Radix” in the market all complied with the relevant regulations under Ophiopogon japonicus (L. f.) Ker-Gawl.. The number of unqualified phloem bundled into microscopic cross-sections. was more than 40%, but the inner cortex cells in the samples showed a comprehensive thickening phenomenon. The results of molecular biology study showed that the four batches of “Ophiopogonis Radix” on the market and the Ophiopogon japonicus (L. f.) Ker-Gawl. were clustered into one, the Liriope spicata (Thunb.) Lour.var. prolifera Y. T. Ma and the Liriope muscari (Decne.) Baily were clustered into one, and showed obvious bar code spacing. The number of phloem bundles was obvious positively correlated with the diameter of wood core by correlation analysis of the number of phloem bundles and the quantitative indexes related to traits.

These four batches of “Ophiopogonis Radix” on the market are the original of genuine products. Combined with the experimental research, it is speculated that the cause of the unqualified number of phloem bundles may be related to the growth years, and the middle column thickness can roughly predict whether it can meet the requirements of pharmacopoeia. To determine whether Ophiopogonis Radix is an accurate medicinal material based on the original, we should not only make a conclusion based on a certain feature of a certain identification method, but also should combine multiple methods to determine accurately.

To investigate an unknown peak at relative retention time (RRT) 0.95 of N-nitrosodiethylamine (NDEA) of irbesartan API with the LC-MS/MS detection method. To elucidate the structure and origin of this unknown peak. Its structure was confirmed with a reference compound, and a method optimization strategy was proposed to eliminate the impact of this unknown peak.

The unknown peak causing interference in the determination of NDEA within sartan drug substances was examined using a full scan method of LC-MS/MS analysis. The unknown peak at RRT 0.95 was validated by conducting high resolution LC-MS analyses and comparing with that of a reference sample. Additionally, a strategy for resolving this issue was proposed in the further investigation.

According to the LC-Q TOF MS results, the unknown peak at RRT 0.95 was confirmed to be 1-pentanamide, a trace regular impurity commonly existing in the sartan samples. As the extract mass of 1-pentanamide was 1 less than that of NDEA, the m/z value of an isotope in 1-pentanamide ([M+H]⁺+1) was 103 and was equal to the m/z value of the main [M+H]⁺ isotope of NDEA under the MRM mode. Furthermore, the [M+H]⁺+1 isotope ion of 1-pentanamide could occur a neutral lose by leaving a molecule of ethylene to generate the isotope ion of m/z 75, which was similar to NDEA in the MRM analysis under ESI positive mode. Therefore, when m/z 103 was used as the precursor ion and m/z 75 was used as the quantitative ion, 1-pentanamide will interfere with detection of NDEA.

In the LC-MS/MS analysis for NDEA, the unknown peak observed at a retention time of 0.95 has been identified as 1-pentanamide. This compound is confirmed as a common trace impurity that regularly occurs during the production of sartan active pharmaceutical ingredients (APIs), rather than being another nitrosamine impurity. Subsequent research has revealed that employing a more selective quantitative model and choosing m/z 47 as the quantifier ion effectively avoid the unknown peak caused by 1-pentanamide.

To establish the research method for the sealing integrity of the injection bottle packaging system for human albumin injection.

For the injection bottle packaging system of human albumin injection, positive control samples with pore diameter of 1 μm were prepared using a glass micropipette, positive control samples with pore diameter of 2, 5 and 10 μm were prepared using laser drilling. Two deterministic sealing integrity testing methods, vacuum decay method and high-voltage leak detection method were developed and tested.

The vacuum decay method could not effectively detect the leakage of the packaging system for the human albumin drug preparation due to the blockage of the leakage hole. In contrast, the high voltage leak detection method effectively avoided undetected leakage caused by the blockage of the leakage hole by the liquid medicine. The test voltage was set at 9 kV, with a threshold of 15 W. Method validation demonstrated that the high voltage leak detection method exhibited good repeatability, intermediate precision, accuracy, and durability, with a leakage detection limit of 1 μm.

The high-voltage leak detection method can serve as an effective means of inspecting the sealing integrity of the injection bottle packaging system for human albumin injection. The procedure is straightforward, and the results are both accurate and reliable, while also being non-destructive to the packaging. This method is well-suited for the sealing integrity inspection of commercial product packaging systems.

To establish a high-performance liquid chromatography quadrupole time-of-flight mass spectrometry (HPLC-Q TOF MS) method for the simultaneous analysis of illegal 4 anti-obesity small molecule drugs and 4 glucagon-like peptide-1 (GLP-1) peptide additives.

The samples were extracted by ultrasound using 50% acetonitrile as the extraction solvent. After centrifugation, the supernatant was taken and separated using an Agilent EC-_C18 chromatographic column (150 mm×3.0 mm, 2.7 μm). Acetonitrile (0.1% formic acid) and water (0.1% formic acid) were used as mobile phases, with gradient elution at a flow rate of 0.3 mL·min^-1 and column temperature of 40 ℃. Adopting positive ion full scanning and target ion secondary fragment scanning methods, with a fragmentation voltage of 150 V. The scanning range of the primary mass spectrometry was m/z 100-3 200, and the scanning range of the secondary mass spectrometry was m/z 50-3 200, with a scanning speed of 1 mass spectrum per second. Establish a data spectral library based on the chromatographic retention time, primary mass spectrometry, and secondary mass spectrometry information of the reference standard, and confirmed the structure through database comparison.

The screening detection limit for peptides was 0.5 µg·mL^-1, while small molecular drugs was 0.05 µg·mL^-1. The recoveries were in the range of 79.4% to 115.8%，with the relative standard deviations of 0.21% to 9.7%. Using this method, 20 batches of anti-obesity drugs were tested, in which semaglutide was identified in 4 samples and sibutratmine was identified in 1 batche.

Compared with the complementary method No. 2012005 by the China Food and Drug Administration, the method established in this study can simultaneously analyze small and large molecules (the relative molecular mass<5 000), featuring high efficiency and accuracy.

To establish an HPLC method for the determination of residual solvent formic acid and cyanoacetic acid content in tofacitinib citrate.

The separation was performed on a Waters Atlantis T3(250 mm×4.6 mm, 5 μm) column with a gradient elution of 0.02 mol·L^-1 potassium dihydrogen phosphate (pH was adjusted to 2.5 with phosphoric acid) (A) - methanol (B) as the mobile phases. The flow rate was 0.5 mL·min^-1 and the column temperature was 35 ℃. Detection wavelength was 210 nm.

Formic acid and cyanoacetic acid were well separated from the adjacent peaks (R>5). The linearity was good in the concentration ranges of 9.930-107.780 μg·mL^-1 and 14.727-98.908 μg·mL^-1 (r = 0.999 9, 0.999 8), respectively. The average recoveries (n=9) were 100.2% and 105.3%, with RSDs of 2.3% and 4.1%, respectively. The determination results of three batches of samples were 0.024%, 0.026%, 0.028% (cyanoacetic acid), respectively.

The established method is proved to be suitable for the determination of formic acid and cyanoacetic acid in tofacitinib citrate.

To discuss the influence of the pharmaceutical excipients on the nitrosamine formation in drugs by studying the nitrate and nitrite in native starch and modified starch.

A methodologyfor the determination of nitrate and nitrite in starch was developed by ion chromatography with a suppressed conductivity detector. A Dionex IonPac anion-exchange column (AS11-HC, 4 mm, 2×250 mm) at a temperature of 30 ℃ was utilized. An electrolytically generated potassium hydroxide solution with a concentration of 0.015 mol·L^-1 was delivered at a rate of 1.0 mL·min^-1. The injection volume was 25 µL. The method was validated in terms of specificity, linearity, LOD & LOQ, precision, accuracy and robustness.

119 samples from 9 different categories of native starch and modified starch were analyzed. Nitrate was detected in all 38 samples of native starch in a range of 1 - 70 µg·g^-1, and identified in 55 out of 81 samples of modified starch in a range of 2 - 1 110 µg·g^-1, Among the modified starch samples, dextrin had the maximum concentration of nitrate. Similarly, nitrite was detected in 33 out of 38 samples of native starch in a range of 0.3 - 2.5 µg·g^-1, and 17 out of 81 samples of modified starch in a range of 1.1 - 13.0 µg·g^-1.

This method has been proven to be suitable for determining the nitrate and nitrite in starch based pharmaceutical excipients. The controlling of nitrate and nitrite in the pharmaceutical excipients is an important part of the nitrosamine risk control in the drugs and should be given sufficient attention.

To collect 37 batches of Scrophulariae Radix decoction pieces from the market as the research object, and to establish a quality grade standard of Scrophulariae Radix decoction pieces based on the correlation between quantitative appearance characteristics and intrinsic quality.

According to the traditional quality evaluation method, 37 batches of Scrophulariae Radix decoction pieces were divided into two grades—A and B. Their appearance characteristics were quantified while their moisture, ash, leaching and intrinsic quality indexes were detected, respectively. Finally, the independent sample t test analysis, orthogonal partial least squares discriminant analysis (OPLS-DA) and Pearson correlation analysis were carried out to reveal the correlation between appearance characteristics and intrinsic quality.

On the basis of meeting the quality control standard of Scrophulariae Radix recorded in the 2020 edition of the Pharmacopoeia of the People's Republic of China（referred to as Chinese Pharmacopoeia, ChP）, it was found that water-soluble extractives and alcohol-soluble extractives were the key indexes for the grades classification of Scrophulariae Radix decoction pieces. In summary, the lower the chromatic value and the greater the density（ρ）of Scrophulariae Radix decoction pieces, the higher the water-soluble extractives and the lower the alcohol-soluble extractives. The grading quality standard of “selected goods” and “gradeless and uniformly priced goods” of Scrophulariae Radix decoction pieces were preliminarily drawn up: the chromatic value（E^*ab）of “selected goods” of Scrophulariae Radix decoction pieces should be less than or equal to 5.5, the ρ of Scrophulariae Radix decoction pieces should be greater than or equal to 1.4 g·cm^-3, and the content of water-soluble extractives should be greater than or equal to 77%, the value of E^*ab of “gradeless and uniformly priced goods” of Scrophulariae Radix decoction pieces should be between 5.5 and 14, the value of ρ should be between 0.8 and 1.4 g·cm^-3, and the content of water-soluble extractives should be greater than or equal to 60%.

The appearance characteristics of Scrophulariae Radix decoction pieces are related with the internal quality, which is consistent with the traditional quality evaluation method, and it is convenient to scientifically judge the internal quality of Scrophulariae Radix decoction pieces from its appearance characteristics.

To establish a method for the identification and characterization of Herba Hyssopi and its adulterants, Nepeta bracteata and Hyssopus officinalis.

Applying stereo microscopy, optical microscopy, and their digital image techniques, the author conducted a comparative study on Herba Hyssopi its adulterants in terms of macroscopic and microscopic characteristics of the whole plant, small samples, and local features (stems, leaves, and flowers), well as the cross sections of stems and leaves and the microscopic characteristics of powders.

There were significant differences in stem, leaf, flower and odor among Hyssopus cuspidatus, Hyssopus officinalis and Nepeta bracteata. The microscopical characteristics of stem cross section were slightly different, but there were obvious differences in leaf cross section and medicinal powder of the three species.

In this study, the pharmacognostic identification characteristics of Herba Hyssopi and its adulterants, Nepeta bracteata and Hyssopus officinalis, were summarized, which can provide the identification basis for the supervision, standard revision and clinical use of Herba Hyssopi.

To establish a multiple fingerprint analysis method for polysaccharides from Sanghuangporus sanghuang, providing a reference for quality evaluation.

High-performance gel filtration chromatography(HPGFC-RID) was used with a TSK-GEL^®G3000 PWXL column (7.8 mm×30 cm, 7 μm). The mobile phase was 20 mmol·L^-1 HAc-NaAc buffer (pH 5.7) at a flow rate of 0.5 mL·min^-1. The injection volume was 15 μL, column temperature was 35 ℃, and detection was carried out using a RID detector. Fourier transform infrared(FT-IR) spectroscopy was performed in the range of 4 000 to 400 cm^-1 with a resolution of 4 cm^-1 and 16 scans. HPLC-UV monosaccharide profiling was done using an Agilent 5 HC-C₁₈ column (250 mm×4.6 mm, 5 μm), with acetonitrile-0.02 mol·L^-1 ammonium acetate (20：80) as the mobile phase, at a flow rate of 1 mL·min^-1, detection wavelength of 250 nm, an injection volume of 10 μL, and column temperature of 35 ℃. An UV detector was used for detection. Multiple fingerprints of Sanghuangporus polysaccharides were established, and the mass average molar mass, characteristic absorptive functional groups, and monosaccharide compositions of 29 batches of Sanghuangporus polysaccharides were compared and analyzed for their intraspecific and interspecific variations by combining with chemometrics and principal component analysis (PCA).

HPGFC-RID monosaccharide profiling revealed that the polysaccharides from Sanghuangporus sanghuang (SH), Sanghuangporus vaninii (YH), Sanghuangporus baumii (BH), and Phellinus pini (SHH) all exhibited two major chromatographic peaks (P1, P2). For P1, BH polysaccharides had the highest molecular weight, followed by SSH, SH, and YH polysaccharides. For P2, YH polysaccharides had the highest molecular weight, followed by SSH, SH, and BH polysaccharides. FT-IR analysis indicated that SH, YH, BH, and SSH polysaccharides shared similar infrared absorption peaks, with no significant differences between species or within species. However, the intensity of the main uptake peaks in the range of 1 800-900 cm^-1 was different between bagged and linden cultivated Sanghuangporus polysaccharides. HPLC-UV monosaccharide profiling combined with chemometrics revealed that Sanghuangporus polysaccharides consist of mannose, rhamnose, glucose, xylose, and fucose. Glucose content was the highest in SH, YH (linden cultivation), BH, and SSH polysaccharides, while xylose was the highest in YH (bagged cultivation). Monosaccharide composition varied significantly among different Sanghuangporus varieties. PCA grouped YH (bagged cultivation) as one class, and SH, YH (linden cultivation), BH, and SSH as another, indicating that cultivation methods also influence Sanghuangporus quality.

The multiple fingerprint analysis method for Sanghuangporus polysaccharides effectively evaluates the quality of different Sanghuangporus varieties, providing a foundation for quality control.

To study the rapid identification of cow-bezoar and its substitutes medicinal herbs using the technique of rapid evaporative ionization mass spectrometry (REIMS) couple with machine learning.

The samples were ionized and determined by REIMS with m/z 50-1 200 as scanning range in sensitive mode and negative ion mode, 0.2 s as scanning time, and using dry burning method. REIMS data of samples was recorded as continuous mode. Then the general situation of REIMS data distribution was studied and analyzed through the methods of cluster analysis and principal component analysis. Some models or algorithms, such as partial least squares discriminant analysis (PLS-DA), logistic regression (LR), decision tree (DT), random forest (RF) and adaptive boosting (AdaBoost, with LR and DT as base estimator respectively) were established. In the models training procedure, simulation synthesis data generated by algorithms of GaussianCopula, CTGAN, CopulaGAN and TVAE joined the original training set data as the new training set.

AdaBoost (DT as base estimator) trained with the new training set was the best model which could accurately predict cow-bezoar and its substitutes medicinal herbs. The accuracy for identifying the test set was 0.97, the precision was 0.90, the recall was 0.97, the F1 score was 0.93, and the AUC of ROC was 1.00. The probability output from the model could also be flexibly used by adjusting the probability threshold according to the actual application scenarios of drug regulation.

The combination of REIMS technology and machine learning technology can achieve fast and accurate recognition of cow-bezoar and its substitutes medicinal herbs.