Natural product-based drug combinations (NPDCs) present distinctive advantages in treating complex diseases. While high-throughput screening (HTS) and conventional computational methods have partially accelerated synergistic drug combination discovery, their applications remain constrained by experimental data fragmentation, high costs, and extensive combinatorial space. Recent developments in artificial intelligence (AI), encompassing traditional machine learning and deep learning algorithms, have been extensively applied in NPDC identification. Through the integration of multi-source heterogeneous data and autonomous feature extraction, prediction accuracy has markedly improved, offering a robust technical approach for novel NPDC discovery. This review comprehensively examines recent advances in AI-driven NPDC prediction, presents relevant data resources and algorithmic frameworks, and evaluates current limitations and future prospects. AI methodologies are anticipated to substantially expedite NPDC discovery and inform experimental validation.
Artificial Intelligence ; Biological Products/chemistry* ; Humans ; Drug Combinations ; Drug Discovery/methods* ; Machine Learning ; Algorithms

Objective:To construct and validate a three-level early warning model of moderate to severe ovarian hyperstimulation syndrome (OHSS) in assisted reproductive technology.Methods:A case-control study was conducted. Totally 10 181 infertility patients who underwent in vitro fertilization treatment in Reproductive Medicine Center, Women and Children's Medical Center Affiliated to Guangzhou Medical University from April 2013 to April 2023 were retrospectively analyzed and divided into modeling group (8 145 cases) and validation group (2 036 cases) by random number table method. The clinical data of the two groups were analyzed, and the risk factors affecting the occurrence of moderate and severe OHSS after oocytes retrieval were screened by multi-factor logistic regression analysis. The early warning model was established and the column diagram was drawn at three nodes which were before ovarian stimulation, before trigger and 3 d after oocyte retrieval. The receiver operating characteristic (ROC) curve and calibration curve were used to verify the models. Results:The antral follicle count (AFC, OR=1.045, 95% CI: 1.020-1.071, P<0.001), anti-Müllerian hormone (AMH)>3.36 μg/L ( OR=7.135, 95% CI: 2.084-24.432, P=0.002) and number of cycles ( OR=0.149, 95% CI: 0.022-1.026, P=0.049) were included in the pre-stimulation prediction model. AFC ( OR=1.046, 95% CI: 1.018-1.074, P=0.001), AMH>3.36 μg/L ( OR=5.780, 95% CI: 1.661-20.116, P=0.006), gonadotropin releasing hormone-agonist protocols ( OR=3.895, 95% CI=1.913-7.931, P<0.001), estrogen peak≥18 350 pmol/L ( OR=2.258, 95% CI: 1.092-4.666, P=0.028), the number of follicles with a diameter of ≥10 mm>20 ( OR=2.377, 95% CI: 1.092-5.172, P=0.029) were included in the pre-trigger prediction model. AMH>3.36 μg/L ( OR=8.374, 95% CI: 2.417-29.019, P=0.001), estrogen peak≥18 350 pmol/L ( OR=3.947, 95% CI: 1.533-10.167, P=0.004), total number of oocytes retrived ( OR=1.042, 95% CI: 0.996-1.090, P=0.025), abdominal distension ( OR=60.181, 95% CI: 22.515-160.854, P<0.001), fresh transplantation ( OR=21.766, 95% CI: 7.119-66.544, P<0.001), human chorionic gonadotropin trigger ( OR=17.752, 95% CI: 3.993-78.924, P<0.001) were included in the prediction model of 3 d after oocyte retrieval. The areas under ROC curves of the three models were 0.830 (95% CI: 0.782-0.878), 0.859 (95% CI: 0.812-0.906) and 0.948 (95% CI: 0.919-0.977), respectively. The areas under ROC curves of the validation groups of the three models were 0.922 (95% CI: 0.880-0.965), 0.936 (95% CI: 0.886-0.986), and 0.971 (95% CI: 0.938-0.999), respectively. The calibration curve indicated that the early-warning evaluation model has good stability. Conclusion:The three-level early warning model of moderate and severe OHSS has good differentiation, reliable predictability and clinical practicability, which is conducive to the dynamic and continuous assessment of the risk of moderate and severe OHSS, adjustment of treatment plan at any time, and timely adoption of effective preventive measures.

Objective:To construct and validate a three-level early warning model of moderate to severe ovarian hyperstimulation syndrome (OHSS) in assisted reproductive technology.Methods:A case-control study was conducted. Totally 10 181 infertility patients who underwent in vitro fertilization treatment in Reproductive Medicine Center, Women and Children's Medical Center Affiliated to Guangzhou Medical University from April 2013 to April 2023 were retrospectively analyzed and divided into modeling group (8 145 cases) and validation group (2 036 cases) by random number table method. The clinical data of the two groups were analyzed, and the risk factors affecting the occurrence of moderate and severe OHSS after oocytes retrieval were screened by multi-factor logistic regression analysis. The early warning model was established and the column diagram was drawn at three nodes which were before ovarian stimulation, before trigger and 3 d after oocyte retrieval. The receiver operating characteristic (ROC) curve and calibration curve were used to verify the models. Results:The antral follicle count (AFC, OR=1.045, 95% CI: 1.020-1.071, P<0.001), anti-Müllerian hormone (AMH)>3.36 μg/L ( OR=7.135, 95% CI: 2.084-24.432, P=0.002) and number of cycles ( OR=0.149, 95% CI: 0.022-1.026, P=0.049) were included in the pre-stimulation prediction model. AFC ( OR=1.046, 95% CI: 1.018-1.074, P=0.001), AMH>3.36 μg/L ( OR=5.780, 95% CI: 1.661-20.116, P=0.006), gonadotropin releasing hormone-agonist protocols ( OR=3.895, 95% CI=1.913-7.931, P<0.001), estrogen peak≥18 350 pmol/L ( OR=2.258, 95% CI: 1.092-4.666, P=0.028), the number of follicles with a diameter of ≥10 mm>20 ( OR=2.377, 95% CI: 1.092-5.172, P=0.029) were included in the pre-trigger prediction model. AMH>3.36 μg/L ( OR=8.374, 95% CI: 2.417-29.019, P=0.001), estrogen peak≥18 350 pmol/L ( OR=3.947, 95% CI: 1.533-10.167, P=0.004), total number of oocytes retrived ( OR=1.042, 95% CI: 0.996-1.090, P=0.025), abdominal distension ( OR=60.181, 95% CI: 22.515-160.854, P<0.001), fresh transplantation ( OR=21.766, 95% CI: 7.119-66.544, P<0.001), human chorionic gonadotropin trigger ( OR=17.752, 95% CI: 3.993-78.924, P<0.001) were included in the prediction model of 3 d after oocyte retrieval. The areas under ROC curves of the three models were 0.830 (95% CI: 0.782-0.878), 0.859 (95% CI: 0.812-0.906) and 0.948 (95% CI: 0.919-0.977), respectively. The areas under ROC curves of the validation groups of the three models were 0.922 (95% CI: 0.880-0.965), 0.936 (95% CI: 0.886-0.986), and 0.971 (95% CI: 0.938-0.999), respectively. The calibration curve indicated that the early-warning evaluation model has good stability. Conclusion:The three-level early warning model of moderate and severe OHSS has good differentiation, reliable predictability and clinical practicability, which is conducive to the dynamic and continuous assessment of the risk of moderate and severe OHSS, adjustment of treatment plan at any time, and timely adoption of effective preventive measures.

Objective:To analyze the assessment tools for quality of life of children with autism spectrum disorder (ASD) at home and abroad, so as to provide a basis for healthcare professionals to select and revise the assessment tools suitable for the quality of life of children with ASD in China.Methods:A scoping review method was used to systematically search the literature on quality of life assessment tools for children with ASD included in PubMed, Embase, Web of Science, Cochrane Library, China National Knowledge Infrastructure, Wanfang Data, China Biology Medicine disc and VIP. The search period was from the establishment of the database to March 12, 2025. Literature that met the inclusion criteria was screened, basic information about the assessment tools was extracted, and the search results were reported in a normalized manner.Results:A total of 3 905 articles were retrieved and 29 articles were finally included. Eleven quality of life assessment scales for children with ASD were included, including two scales developed to characterize children with ASD.Conclusions:There is a wide variety of quality of life assessment tools for children with ASD, with good overall reliability and validity, but no standardized indigenous scales are developed in China. Existing quality of life assessment tools for children with ASD should be sinicized and improved, and localized assessment tools should be developed.

Objective:To analyze the assessment tools for quality of life of children with autism spectrum disorder (ASD) at home and abroad, so as to provide a basis for healthcare professionals to select and revise the assessment tools suitable for the quality of life of children with ASD in China.Methods:A scoping review method was used to systematically search the literature on quality of life assessment tools for children with ASD included in PubMed, Embase, Web of Science, Cochrane Library, China National Knowledge Infrastructure, Wanfang Data, China Biology Medicine disc and VIP. The search period was from the establishment of the database to March 12, 2025. Literature that met the inclusion criteria was screened, basic information about the assessment tools was extracted, and the search results were reported in a normalized manner.Results:A total of 3 905 articles were retrieved and 29 articles were finally included. Eleven quality of life assessment scales for children with ASD were included, including two scales developed to characterize children with ASD.Conclusions:There is a wide variety of quality of life assessment tools for children with ASD, with good overall reliability and validity, but no standardized indigenous scales are developed in China. Existing quality of life assessment tools for children with ASD should be sinicized and improved, and localized assessment tools should be developed.

Objective:To explore the current status of medication adherence to immunosuppressants in lung transplant recipients and to analyze its correlation with medication beliefs and perceived social support.Methods:This was a cross-sectional study. From June 2022 to April 2023, totally 233 lung transplant recipients who were followed up in the Department of Lung Transplantation at China-Japan Friendship Hospital were selected by convenience sampling. The Basel Assessment of Adherence to Immunosuppressive Medications Scale (BAASIS), the Chinese version of the Beliefs about Medicines Questionnaire, and the Perceived Social Support Scale (PSSS) were used for the survey. Multiple linear regression analysis was conducted to explore the correlation between medication adherence to immunosuppressants, medication beliefs, and perceived social support in these lung transplant recipients.Results:A total of 233 questionnaires were distributed, with 213 valid responses received (91.42%). The incidence of non-adherence to immunosuppressants among the 213 transplant recipients was 41.78% (89/213), with the most common issue being not taking medication on time (27.23%, 58/213). Multiple linear regression analysis showed that age and perceived social support were influencing factors of medication adherence ( P<0.05) . Conclusions:The current level of medication adherence to immunosuppressants in lung transplant recipients is relatively poor. Medication adherence is correlated with age and the level of perceived social support. Healthcare providers should pay attention to medication adherence in younger patients and enhance their perceived social support to increase adherence to immunosuppressive medications.

Objective:To analyze the incidence and clinical phenotype of the concomitant extragenital malformations in the patients with female reproductive tract anomalies.Methods:A retrospective study was conducted using clinical data of hospitalized patients diagnosed with uterine, cervical, or vaginal malformations from January 2003 to December 2022 in Peking Union Medical College Hospital. The malformations were classified according to American Society for Reproductive Medicine müllerian anomalies classification 2021, and in each type, the incidence and specific manifestations of concomitant extragnital malformations were analyzed.Results:A total of 444 patients were included. The overall incidence of concomitant extragenital malformations was 43.5% (193/444), including urinary system, skeletal system, and other system malformations. Renal malformations on the obstructed side were present in all patients with oblique vaginal septum syndrome (100.0%, 78/78). The total incidence of concomitant extragnital malformations was as high as 8/11 in uterus didelphys, 43.5% (10/23) in unicornuate uterus, 33.6% (79/235) in Mayer-Rokitansky-Küster-Hauser syndrome, 18.8% (6/32) in septate uterus and 18.5% (12/65) in cervical agenesis. Urinary system malformations (30.6%, 136/444) and skeletal system malformations (13.5%, 60/444) were the most common concomitant malformations in all types, in which, unilateral renal agenesis and scoliosis were the most common.Conclusions:Urinary and skeletal system malformations are important features of female reproductive tract anomalies. Urologic ultrasonography and spinal roentgenogram are recommended for all patients with female reproductive tract anomalies.

OBJECTIVE: To explore the genetic basis for two children with sporadic neurofibromatosis type 1 (NF1) complicated with nephrotic syndrome (NS). METHODS: Clinical data of the children were collected. Both children were subjected to high-throughput sequencing, and candidate variants were verified by Sanger sequencing. RESULTS: Both children had café-au-lait macules, subaxillary freckle and Lisch nodules. Child 1 also had congenital tibiofibular pseudarthrosis on the left side. Genetic testing revealed that child 1 has harbored a heterozygous c.844C>T variant in the exon 8 of the NF1 gene, whilst child 2 has harbored a heterozygous c.1246C>T variant in the exon 11 of the NF1 gene. Both children were diagnosed with NF1 and have developed pronounced proteinuria, hypoalbuminemia, hypercholesterolemia and pitting edema at the ages of 3 and 10, respectively. Renal biopsy of child 2 has revealed minimal change nephropathy, and the diagnosis of nephrotic syndrome was established. Child 1 was treated with glucocorticoid, and child 2 was treated with glucocorticoid in combination with mycophenolate mofetil. The NS was relieved with no recurrence during 1 year's follow-up. CONCLUSION NF1 combined with NS is rare in the clinical settings. The prognosis of children with NF1 combined with minimal change nephropathy is relatively good. Detection of NF1 gene variant can facilitate early identification and diagnosis of NF1.
Child ; Humans ; Neurofibromatosis 1/genetics* ; Nephrotic Syndrome/genetics* ; Nephrosis, Lipoid ; Glucocorticoids ; Genetic Testing

OBJECTIVE: To explore the influencing factors of depressive symptom of employees in a mould-producing enterprises using two occupational stress models. METHODS: A total of 457 employees in a mould-producing enterprise were selected as study subjects using cluster sampling method. The Center for Epidemiological Study-Depression Scale, the Chinese version of Job Demand Content Questionnaire and the Effort-Reward Imbalance Questionnaire were used to investigate the level of depressive symptom and occupational stress by the job demand control(JDC) model and effort reward imbalance(ERI) model. RESULTS: Among the 457 employees, the median score of depressive symptom was 11.00, the detection rate of depressive symptom in the subjects was 23.4%(107/457). The incidence of occupational stress by the JDC and ERI models was 78.3%(358/457) and 62.6%(286/457), respectively. The rate of depressive symptom was statistically higher in the high occupational stress group than that of the low occupational stress group by the ERI model(26.6% vs 18.1%, P<0.05). The multivariate logistic regression analysis showed that both marital status and monthly income were influencing factors of depressive symptom by the JDC model(P<0.05). The marital status, monthly income and occupational stress were influencing factors of depressive symptom by the ERI model(P<0.05). The risk of depressive symptom was statistically higher in the high occupational stress group than that of the low occupational stress group(P<0.05). CONCLUSION: Employees in mould-producing enterprise have some depressive symptom. The marital status, monthly income and occupational stress by the ERI model are the main influencing factors of depressive symptom.

OBJECTIVE：To stud y the effects of different processed products of Whitmania pigra on hemorheology and coagulation indexes in acute blood stasis model rats. METHODS ：SD rats were randomly divided into blank group ，model group ， aspirin group ，W. pigra hang-dried product low- ，medium- and high-dose groups ，W. pigra talcum powder-ironed product low- ， medium- and high-dose groups ，W. pigra wine bran-processed product low- ，medium- and high-dose groups ，with 6 rats in each group. Except for blank group ，other groups received subcutaneous injection of epinephrine hydrochloride and ice water bath for 15 d to induce acute blood stasis model. From the 8th day of modeling ，rats in aspirin group were given aspirin 0.2 g/kg intragastrically. Rats in each dose group of W. pigra processed products were given relevant medicine 0.35，1.4，3.5 g/kg intragastrically（calculated by crude drug ）. Rats in blank group and model group were given constant volume of normal saline intragastrically， once a day ， for consecutive 8 days. Hemorheology indexes as whole blood viscosity （high， medium and low shearrate ），plasma viscosity ，erythrocyte com deformation index ，erythrocyte aggregation index ，hematocrit， and blood coagulation indexes as prothrombin time （PT）， mail：wcl19960125@163.com activated partial prothrombin time （APTT），thrombin time （TT）were determined. RESU LTS：Compared with blank group ，whole blood viscosity under different shear rates ，plasma viscosity ， erythrocyte aggregation index and hematocrit of model group were increased significantly ，while erythrocyte deformation index was significantly decreased ，PT，TT and APTT were significantly shortened （P＜0.01）. Compared with model group ，whole blood viscosity under different shear rates ，plasma viscosity ，erythrocyte aggregation index and hematocrit of aspirin group and W. pigra hang-dried product ，talcum powder-ironed product ，wine bran-processed product high-dose groups were decreased significantly ， while erythrocyte deformation index were significantly increased ，and PT （only W. pigra talcum powder-ironed products high-dose group），APTT（except for W. pigra hang-dried products high-dose group ）and TT were prolonged significantly. The whole blood viscosity of W. pigra hang-dried product medium-dose group under low shear rate ，and those of W. pigra talcum powder-ironed product low-dose ，wine bran-processed product medium-dose groups under low and medium shear rates were decreased significantly. Erythrocyte deformation index of W. pigra talcum powder-ironed product medium-dose group was increased significantly ，while erythrocyte aggregation index was decreased significantly ，and PT ，TT were prolonged significantly. APTT of W. pigra hang-dried product medium-dose group was prolonged significantly. Hematocrit of W. pigra wine bran-processed product low-dose group was decreased significantly （P＜0.05 or P＜0.01）. CONCLUSIONS ： W. pigra hang-dried， talcum powder-ironed and wine bran-processed product can effectively improve hemorheology indexes and prolong blood coagulation time.