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

The αPD-L1 antibody-based immune checkpoint blockade therapy is still limited by the poor clinical response rate as it is mainly utilized to block surface PD-L1 on tumor cells while ignoring abundant PD-L1 exosomes secreted in the environment, causing tumor immune evasion. Here, we proposed an exosome biogenesis inhibition strategy to suppress tumor exosomes secretion from the source, reducing the inhibitory effect on T cells and enhancing chemo-immunotherapy efficacy. We developed sulfafurazole homodimers (SAS) with disulfide linkages, effectively releasing the drug in response to glutathione (GSH) and inhibiting 4T1 tumor-derived exosomes secretion. Subsequently, gemcitabine (Gem) was encapsulated to induce immunogenic cell death (ICD). Consequently, Gem@SAS inhibited the secretion of tumor exosomes by more than 70%, increased proliferation and granzyme B secretion ability of T cells by more than 2 times, and showed superior efficacy in breast cancer treatment as well as lung metastasis of breast cancer.

Objective:To determine the motion detection uncertainty of the real-time CybeRay ? imaging system and patient intrafractional motion with thermoplastic mask-based immobilization. Methods:Real-time CybeRay ? imaging system was used for irradiation and treatment for head phantom and patients with brain tumors. All patients were immobilized with thermoplastic masks. Real-time imaging was delivered using kilovoltage projection images during radiotherapy. The detected patient motion data was collected from 5 head phantom measurements and 27 treatment fractions of 9 brain tumor patients admitted to Kaifeng Cancer Hospital. The accuracy and uncertainty of the motion monitoring system were determined. Results:The mean and standard deviation (SD) of the detected motion in the X, Y, and Z directions for phantom were (-0.02±0.41) mm, (-0.05±0.22) mm and (0.01±0.35) mm, respectively. The detected motion in the X, Y and Z directions for patents were (-0.13±0.48) mm, (-0.05±0.48) mm and (0.11±0.36) mm, respectively. After removing the motion detection uncertainty, the actual intrafractional motion of patients were (-0.11±0.25) mm, (0±0.43) mm and (0.10±0.08) mm in three directions, respectively. Conclusions:The uncertainty of real-time imaging-based motion monitoring system of CybeRay ? is less than 0.5 mm. It is feasible to apply thermoplastic masks for brain tumor patients in clinical practice, which can provide steady immobilization and limit the SD of patient intrafractional motion within 0.5 mm. Real-time imaging-based motion monitoring system of CybeRay ? is accurate for patient motion monitoring during frameless stereotactic radiosurgery/radiotherapy.

Objective:To investigate the echocardiographic features, consistency of diagnosis between fetal and postnatal periods and postnatal clinical outcomes of fetal pulmonary valve stenosis (PS) with different degrees.Methods:This study was a retrospective cohort study comprising 108 cases of fetal PS diagnosed during the fetal period and followed up postnatally at Xinhua Hospital, Shanghai Jiaotong University School of Medicine from November 2012 to February 2023. Echocardiographic characteristics, including morphological and hemodynamic features were collected for all fetuses who were then were followed up to at least 6 months after birth. One-way analysis of variance and Kruskal-Wallis test were used to compare the differences in the echocardiographic features among fetuses with different degrees of PS. Subsequently, McNemar test was used to assess the consistency of diagnosis between the fetal and postnatal periods. Furthermore, Logistic regression analysis was applied to explore the risk factors for neonatal intervention in fetuses with moderate PS and the receiver operating characteristic (ROC) curve was utilized to ascertain the optimal cut-off value for continuous variables.Results:The age of the mothers of the 108 fetuses at the initial assessment was (30.8±4.0) years, and the gestational age was 26.5 (24.6, 30.0) weeks. The fetuses were categorized into mild (17 cases), moderate (49 cases), and severe groups (42 cases) based on the initial echocardiographic features. Mild PS was characterized by valve thickening and hyperechogenicity combined with systolic flow acceleration or dilation of main pulmonary artery. Moderate PS exhibited both restricted valve motion and a colorful blood flow pattern at the valve orifice. The peak flow velocities of fetuses with moderate and critical PS were notably higher than those in the mild group ((2.66±0.86) and (2.77±1.30) vs. (1.43±0.59)m/s, F=14.52, P<0.001). In critical PS, all cases showed retrograde ductal flow, with a significantly higher proportion of a small right ventricle compared to the mild and moderate PS (42.9% (18/42) vs. 0 and 2.0% (1/49), χ2=31.73, P<0.001). The proportion of severe tricuspid regurgitation was also higher (35.7% (15/42) vs. 0 and 10.2% (5/49), χ2=36.94, P<0.001). Compared to mild and severe PS, the consistency of diagnosis between fetal and postnatal periods in moderate PS was lower (40.8% (20/49) vs.13/17 and 80.3% (35/42), χ 2=12.45, P=0.006). The systolic flow velocity was identified as an independent risk factor for neonatal intervention in fetuses with moderate PS ( OR=7.21, 95% CI2.11-24.62). A flow velocity of ≥2.18 m/s in second trimester and ≥3.15 m/s in third trimester indicated the necessity of neonatal intervention for fetal moderate PS. Among the 108 fetuses, 68 underwent surgical intervention and all survived. Additionally, 39 fetuses were regularly followed up. A sole non-surgical fatality occurred, leading to a 6-month survival rate of 99.1% (107/108). Conclusions:Various degrees of fetal PS demonstrate distinctive morphological and hemodynamic alterations in echocardiography. The disparity in severity between the postnatal and fetal stages requires ongoing monitoring for fetal PS. The prognosis for fetal PS is generally favorable.

Humans ; Amyloidosis ; Amyloid ; Cardiomyopathies

Non-small cell lung cancer (NSCLC) is one of the most common types of cancer in the world and is an important cause for cancer death. Although the application of immunotherapy in recent years has greatly improved the prognosis of NSCLC, there are still huge challenges in the treatment of NSCLC. The immune microenvironment plays an important role in the process of NSCLC development, infiltration and metastasis, and they can interact and influence each other, forming a vicious circle. Notably, single-cell RNA sequencing enables high-resolution analysis of individual cells and is of great value in revealing cell types, cell evolution trajectories, molecular mechanisms of cell differentiation, and intercellular regulation within the immune microenvironment. Single-cell RNA sequencing is expected to uncover more promising immunotherapies. This article reviews the important researches and latest achievements of single-cell RNA sequencing in the immune microenvironment of NSCLC, and aims to explore the significance of applying single-cell RNA sequencing to analyze the immune microenvironment of NSCLC.

Lung cancer is a malignant tumor with the highest mortality worldwide, and its early diagnosis and evaluation have a crucial impact on the comprehensive treatment of patients. Early preoperative diagnosis of lung cancer depends on a variety of imaging and tumor marker indicators, but it cannot be accurately assessed due to its high false positive rate. Liquid biopsy biomarkers can detect circulating tumor cells and DNA in peripheral blood by non-invasive methods and are gradually becoming a powerful diagnostic tool in the field of precision medicine for tumors. This article reviews the research progress of liquid biopsy biomarkers and their combination with clinical imaging features in the early diagnosis of lung cancer.

Objective:To evaluate the clinical radiological features combined with circulating tumor cells in the diagnosis of benign and malignant pulmonary solid nodules.Methods:Clinical data of 437 patients from Shanghai Pulmonary Hospital(SPH cohort) from January to April 2021 and 82 patients from Lanzhou University First Hospital (LZH cohort) from August 2019 to May 2022 were retrospectively included. Patients in Shanghai pulmonary hospital were randomly divided into training set and internal validation set in a ratio of 4∶1 by random number table method and patients in Lanzhou University First Hospital were as external validation set. Independent risk factors were selected by regression analysis of training set constructed a Nomogram prediction model. The performance of the Nomogram prediction model was estimated by applying receiver operating curve( ROC) analysis, tested in different nodules size and intermediate risk IPSNs and tested by calibration curve. Results:Independent risk factors selected by regression analysis for solid pulmonary nodules were age, the level of CTC, pleural Indentation, lobulation, spiculation. The Nomogram prediction mode provided an area under ROC( AUC) of 0.888, 0.833 in internal validation set and external validation set, outperforming radiological features model(0.835, P=0.007; 0.804, P=0.043) Mayo clinical model(0.781, P=0.019; 0.726, P=0.033) and CTCs(0.699, P=0.002; 0.648, P=0.012) in both two validation sets, C-index of 0.888, 0.871 and corrected C-index of 0.853, 0.842 in both two validation sets . The AUC of the prediction model with internal validation set was 0.905 and 0.871 for nodule diameter of 5-20 mm and intermediate risk probability. Conclusion:The prediction model in this study has better diagnostic value and practicability, and is more effective in clinical diagnosis of diseases.

Objective    To establish a machine learning model based on computed tomography (CT) radiomics for preoperatively predicting invasive degree of lung ground-glass nodules (GGNs). Methods    We retrospectively analyzed the clinical data of GGNs patients whose solid component less than 3 cm in the Department of Thoracic Surgery of Shanghai Pulmonary Hospital from March 2021 to July 2021 and the First Hospital of Lanzhou University from January 2019 to May 2022. The lesions were divided into pre-invasiveness and invasiveness according to postoperative pathological results, and the patients were randomly divided into a training set and a test set in a ratio of 7∶3. Radiomic features (1 317) were extracted from CT images of each patient, the max-relevance and min-redundancy (mRMR) was used to screen the top 100 features with the most relevant categories, least absolute shrinkage and selection operator (LASSO) was used to select radiomic features, and the support vector machine (SVM) classifier was used to establish the prediction model. We calculated the area under the curve (AUC), sensitivity, specificity, accuracy, negative predictive value, positive predictive value to evaluate the performance of the model, drawing calibration and decision curves of the prediction model to evaluate the accuracy and clinical benefit of the model, analyzed the performance in the training set and subgroups with different nodule diameters, and compared the prediction performance of this model with Mayo and Brock models. Two primary thoracic surgeons were required to evaluate the invasiveness of GGNs to investigate the clinical utility of the mode. Results    A total of 400 patients were divided into the training set (n=280) and the test set (n=120) according to the admission criteria. There were 267 females and 133 males with an ……

Fibrosis is one of the key factors that lead to the immune exclusion of solid tumors. Although degradation of fiber is a promising strategy, its application was still bottlenecked by the side effects of causing metastasis, resulting in the failure of immunotherapy. Here, we developed an antimetastatic polymer (HPA) for the delivery of chemo-drug and antifibrotic siPAI-1 to form the nano-permeator. Nano-permeator shrank after protonation and deeply penetrated into the tumor core to down-regulate the expression of PAI-1 for antifibrosis, and further promoted the sustained infiltration and activation of T cells for killing tumor cells. Moreover, metastasis after fiber elimination was prevented by multivalent CXCR4 antagonistic HPA to reduce the attraction of CXCL12 secreted by distant organs. The administration of stroma-alleviated immunotherapy increased the infiltration of CD8⁺ T cells to 52.5% in tumor tissues, inhibiting nearly 90% metastasis by HPA in distant organs. The nano-permeator reveals the mechanism and correlation between antifibrosis and antimetastasis and was believed to be the optimizing immunotherapy for solid fibrotic tumors.