To investigate the microbial contamination in commercially available Angelicae Sinensis Radix decoction pieces and analyze the potential safety risks.

According to the General Principles of the Pharmacopoeia of the People’s Republic of China (referred to as the “Chinese Pharmacopoeia”) (2020 edition <1108 Microbial Limit Examination of Chinese Herbal Medicine>)，the total aerobic microbial count (TAMC)，total combined yeasts and molds count (TYMC)，heat-resistant bacterial count，control pathogens，and bile-tolerant gram-negative bacterial count were determined. High-throughput sequencing technology was utilized to analyze the dominant microbial species present in representative samples and identify potential objectionable microorganisms.

There was a high level of uncertainty regarding microbial contamination in commercially available Angelicae Sinensis Radix，and the dominant group in different types had significant differences. The TAMC ranged from 10² to 10⁶ CFU·g^-1，while a uneven contamination of bile-tolerant gram-negative bacteria，including potentially pathogenic species such as Acinetobacter baumannii，Klebsiella pneumoniae and Enterobacter cloacae，but Salmonella was not detected in any samples. The total combined yeasts and molds count ranged from 10¹ to 10³ CFU·g^-1.

The microbial contamination in Angelicae Sinensis Radix is severe and uneven. More data is needed to improve testing methods and evaluation standards for assessing the risk of microbial contamination in Chinese herbal medicine. High-throughput sequencing can provide a more accurate assessment of microbial risks in decoction samples of traditional Chinese herbal medicine.

To provide reference for quality control and authenticity identification of Arnebiae Radix medicinal materials and decoction pieces in the market. By studied on the ITS2 sequences’ characters of imported Arnebiae Radix，based on DNA barcoding and PCR-RFLP technologies.

The ITS2 region was selected as the DNA barcode sequence for comparison and identification of imported Arnebiae Radix and reference medicinal materials. The ITS2 sequences of imported Arnebiae Radix from different sources with reference medicinal materials were compared based on DNA barcoding and PCR-RFLP technologies.

After the restriction endonucliase AluI enzyme digestion，the agarose-gel electrophoresis results of 39 imported Arnebiae Radix samples showed that，only DH3 had bands at around 500 bp，and none bands between 100 bp and 300 bp. And the results of other imported Arnebiae Radix samples had two or three obvious bands between 100 bp and 300 bp. The ITS2 sequences of imported Arnebiae Radix samples were compared with the reference medicinal materials，among which DH3 had the largest differences of 15 bases compared to the reference medicinal materials，the ITS2 sequence of F2 was same to the reference medicinal materials，and other imported Arnebiae Radix samples had 1-9 bases difference compared to the reference medicinal materials. The clustering results showed that the imported Arnebiae Radix sample DH3 was clearly distinguished from other imported Arnebiae Radix samples and reference medicinal materials which was in a single branch. There were 14 samples，which were clustered together with the reference medicinal materials in one branch with support rate ≥50%.

The ITS2 region is selected to compare the similarities and differences of ITS2 sequences between imported Arnebiae Radix samples and reference medicinal materials based on DNA barcode and PCR-RFLP technologies，which provids a reference for effective identification of Arnebiae Radix medicinal materials and decoction pieces，and a strong guarantee for market supervision of Arnebiae Radix medicinal materials.

To establish a three classification model for cultivated，semi-wild，and wild Astragali Radix characterized by flavonoids，and explore and evaluate the application of techniques of automated machine learning and data augmentation in the field of drug analysis.

Firstly，correlation analysis and principal component analysis were conducted on the flavonoid content data of Astragali Radix，and models of decision tree and logistic regression were established to analyze the importance of flavonoid components based on the models. Then，using the AutoGluon framework with 5 as num_bag_folds，2 sets of 30 models respectively through 64 batches of real data and 600 batches of virtual data generated based on real data with the TVAE table data generation algorithm for training were obtained，and these models were evaluated by accuracy.

The analysis of machine learning models，indicated that formononetin，campanulin and onospin played the important roles in the quality control of Astragali Radix，especially for the source grade control. The accuracy of model prediction showed that the models based on Neural Net and tree-model always had the best classification effect for Astragali Radix. The virtual data generated by data augmentation technique is basically consistent with the actual data in terms of the accuracy trend of the model training process.

Related techniques of machine learning have good application value in the classification of Astragali Radix characterized by flavonoids.

To compare the regulatory effects of authentic and counterfeit Arnebiae Radix on intestinal flora in mice based on metagenomic sequencing.

Firstly，24 clean grade female BLAB/C mice were randomly divided into 3 groups：blank control group，A1 (Arnebia euchroma) group and A2 (Arnebiae Radix whose origins were not included in Chinese Pharmacopoeia) group. After gavage arrived at the specified time，colon contents (feces)，ileal contents (feces) and small intestinal contents (feces，except ileal parts) were extracted for intestinal flora analysis. Genomic DNA was extracted and amplified by PCR from the extracted mouse intestinal contents. The PCR products were mixed and purified. Then library was constructed and sequenced. After quality control of sequencing data and removal of chimera sequence，the final effective data was obtained. Operational taxonomic unit(OTU) clustering and species annotation were performed on the obtained valid data，and sample diversity analysis was conducted.

In this study，both A1 (Arnebia euchroma) group and A2 (Arnebiae Radix non-pharmacopoeia) group reduced the diversity of mice colon microbiota. At the phylum level，group A1 significantly increased the abundance of Firmicutes in the small intestine and ileum，and groups A1 and A2 significantly increased the relative abundance of Bacteroidetes in the colon. At the genus level，group A1 significantly increased the relative abundance of Lactobacillus in the small intestine of mice，and group A2 significantly increased the relative abundance of Lactobacillus in the ileum of mice. Group A1 increased the relative abundance of Lactobacillus，one of the dominant bacteria in the colon of mice，and group A2 increased the relative abundance of Bacteroides. Alistipes mainly existed in the colon，A2 group significantly reduced the relative abundance of Alistipes in the colon of mice，while Alistipes in the A1 group was cultivated.

According to the results of the regulation effect of intestinal flora，the intestinal flora regulation effect in the two experimental groups of the authentic Arnebiae Radix and its confusion products from markets are not consistent，the results of this study can provide a theoretical basis for further exploring the mechanism of Arnebiae Radix.

To establish a dual-wavelength method for the determination of amylose and amylopectin in potato starch in Tibet.

The determination wavelengths of amylopectin and amylopectin in Tibetan potato were 623 and 557 nm，respectively，and the reference wavelengths were 498 and 729 nm. The contents in 11 batches of starch samples collected were determined and analyzed.

The results showed that the mass concentration of amylose and amylopectin showed a good linear relationship (r=0.999 9，r=0.999 8) in the range of 0- 66 μg·mL^-1 and 0-90 μg·mL^-1，respectively. The average recoveries of amylose and amylopectin were 102.8% (RSD=1.8%) and 98.5% (RSD=2.2%). The precision (RSD=0.071%，RSD=0.31%)，repeatability (RSD=0.26%，RSD=2.4%) and stability (RSD=0.14%，RSD=1.4%) were all in compliance with regulations. The contents of amylose and amylopectin in Tibetan potato starch were 36.74% and 45.87%，respectively. Compared with the commercially available potato starch，the amylose content of Tibetan potato starch was significantly increased (P<0.001)，while the amylopectin content was significantly decreased (P<0.05).

The method is proved to be suitable for the quality evaluation of potato starch in Tibet. There are significant differences in the contents of amylopectin and amylopectin between potato starch in Tibet and potato starch in the market，which provides an experimental basis for the establishment of quality standards of potato starch in Tibet and the development of high-value products.

To establish a ultra-high performance liquid chromatography-mass spectrum in series method for the quantification of 7-hydroxycoumarin，fraxetin，isofraxidin and scopoletin in the aqueous extract of Acanthopanacis Senticosi Radix et Rhizoma Seu Caulis in rat plasma，and to evaluate the pharmacokinetic behavior and bioavailability in rats.

The chromatography was performed on Thermo Scientific Hypersil GOLD aQ（100 mm×2.1 mm，1.9 μm）column with 0.1% aqueous formic acid -0.1% acetonitrile formic acid solution (82：18) as the mobile phase at the flow rate of 0.30 mL·min^-1，and column temperature of 40 ℃. Mass spectrometry positive ion mode scan was used，and sample size was 5 μL. Healthy SD rats were selected for a single gavage of 10 mL·kg^-1 of Acanthopanacis Senticosi Radix et Rhizoma Seu Caulis water extract (equivalent to the dose of 1 g·kg^-1 of the original drug). Plasma concentration of the substances was determined at different time intervals after administration，and the pharmacokinetic parameters were calculated using DAS software by non-atrioventricular model fitting.

The methodological results showed that 7-hydroxycoumarin (r=0.999 4)，fraxetin (r=0.998 9)，isofraxidin (r=0.999 3) and scopoletin (r=0.998 4) had good linearity in the range of 0.05-55 μg·mL^-1，and RSDs of precision of the four substances were all less than 15%. The recovery was 85%-115%. The absolute bioavailability ranged from 59% to 78%，and the relative bioavailability ranged from 80% to 87%. The extraction recovery，matrix effect and stability met the relevant requirements.

After a single gavage of the aqueous extract of Acanthopanacis Senticosi Radix et Rhizoma Seu Caulis in rats，the four substances to be tested are absorbed and eliminated by rats. The method validation results are in line with the guiding principles of biological sample analysis methods，and can be applied to evaluate the pharmacokinetic behavior and bioavailability of the aqueous extract of Acanthopanacis Senticosi Radix et Rhizoma Seu Caulis in rats.

To compare the quality of wild and cultivated Arnebiae Radix，using macroscopic investigation and chemometric analysis of the different components in wild and cultivated Arnebiae Radix from three different habitats.

Wild and cultivated Arnebiae Radix were collected and their macroscopic features were compared. Using the ACQUITY UPLC BEH C₁₈ (2.1 mm×100 mm，1.7 μm) column，with acetonitrile-0.05% formic acid water as the mobile phase，the contents of D-shikonin，acetylshikonin，β-acetoxyisovalerylshikonin，isobutyrylshikonin，β，β’-dimethylacrylalkannin and isovalerylshikonin in 48 batches of wild and cultivated Arnebiae Radix were determined. The detection wavelength was 275 nm and the flow rate was 0.2 mL·min^-1. PCA and OPLS-DA were performed to reveal the differential components of wild and cultivated Arnebiae Radix.

There were great differences in macroscopic features of wild and cultivated Arnebiae Radix，and the linear relationship between the contents of six naphthoquinone components was good. The correlation coefficients were above 0.999，the average recovery rates were 93.4%-102.9%，and the RSDs were less than 3.0%. The contents of six components in different batches of wild and cultivated Arnebiae Radix were quite different，and the contents of D-shikonin and acetylshikonin in wild products were significantly higher than those in cultivated products，indicating that there were still certain differences between wild products and cultivated products. The PCA model established could distinguish wild products and cultivars，and two differentiating components in wild products and cultivars were revealed by OPLS-DA，namely isobutyryl shikonin，β，β’-dimethylacrylalkannin.

By comparing the core size，cork curl degree and specific odor of wild and cultivated products，the two can be identified. The established content determination method is repeatable，specific，stable and feasible. The differential components in wild and cultivated Arnebiae Radix in three different regions are identified，which provides a basis for the quality control of Arnebiae Radix and provides ideas for expanding the source of Arnebiae Radix.

To establish a gas chromatography-mass spectrometry (GC-MS) method for the determination of 16 photoinitiator residuce levels in medicinal composite membranes，including benzophenone (BP) and 4-methylbenzophenone (4-MBP).

The samples were extracted by ethylacetate，separated with TR-5MS column. The mass spectrum was analyzed with EI ionization，positive ion mode and selective reaction monitoring (SRM) mode and quantified with internal standard method.

The 16 kinds of photoinitiators had good linear relationships in the range of 0.05-2.0 μg·mL^-1，and the correlation coefficients (r) were more than 0.997. The limits of detection (LODs) were 0.03 μg·mL^-1. The average recoveries were (n=6) ranged from 81.3% to 94.0%，with RSDs of 2.1%-6.2%. The detection rate of 4-Isopropylthioxanthone (4-ITX) in the medicinal composite membranes was 30%，and the highest content was 0.02 μg·cm^-2.

The method is accurate with high sensitivity and simple pretreatment，which can be used for the detecting the kinds of photoinitiators in medicinal composite membranes.