Objective:This study aimed to construct a prediction model for diagnosis of proliferative lupus nephritis based on a machine learning algorithm. Additionally, a user-friendly platform was developed to propose a non-invasive method to assist the pathologic classification of lupus nephritis.Methods:A retrospective analysis was conducted on clinical and pathological data of lupus nephritis patients confirmed by renal biopsy at Zhengzhou University People′s Hospital from January 2017 to August 2023. The study population was randomly divided into training and testing sets in a 7∶3 ratio. Utilizing six machine learning algorithms, classification models were developed. The predictive performance of each model was assessed using metrics such as accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC). The optimal model, once identified, was deployed as a web-based calculator for convenient model application. SPSS 25.0 and R 4.2.2 were used to analyze the data.Results:The study included a total of 212 patients, with 138 cases with proliferative lupus nephritis and 74 cases with non-proliferative lupus nephritis. The AUC values for the six models, namely logistic regression, decision tree, random forest, support vector machine, extreme gradient boosting, and light gradient boosting machine, were 0.79, 0.62, 0.79, 0.88, 0.81, and 0.77, respectively; the accuracy rates were 82.54%, 65.08%, 74.60%, 85.71%, 69.84%, 71.43%, respectively. Among them, the support vector machine model demonstrated the optimal performance. This model had deployed as a web-based calculator. Based on feature importance scores, the top 10 influencing factors were identified, including anti URNP antibody, immunoglobulin G, serum globulin, estimated glomerular filtration rate, anti Smith antibody, BMI index, anti dsDNA antibody, uric acid, anti-Rib.p antibody, and gender.Conclusion:A prediction model based on machine learning algorithms was successfully established, and a web calculator was developed to offer a simple and non-invasive method for diagnosing proliferative lupus nephritis. This can assist clinicians in evaluating the risk-benefit ratio of kidney biopsy in patients with lupus nephritis.

BACKGROUND: Cervical squamous cell carcinoma (CESC), the most common subtype of cervical cancer, is primarily caused by the high-risk human papillomavirus (HPV) infection and genetic susceptibility. Single-cell RNA sequencing (scRNA-seq) has been widely used in CESC research to uncover the diversity of cell types and states within tumor tissues, enabling a detailed study of the tumor microenvironment (TME). This technology allows precise mapping of HPV infection in cervical tissues, providing valuable insights into the initiation and progression of HPV-mediated malignant transformation. METHODS: We performed the scRNA-seq to characterize gene expression in tumor tissues and paired adjacent para-cancerous tissues from four patients with early-stage CESC using the 10× Genomics platform. The HPV infection and its subtypes were identified using the scRNA data and viral sequence mapping, and trajectory analyses were performed using HPV+ or HPV- cells. Interactions between different types of keratinized cells and their interactions with other cell types were identified, and pathways and specificity markers were screened for proliferating keratinized cells. The Cancer Genome Atlas (TCGA) dataset was used to verify the prognostic correlation between tumor-specific PI3+S100A7+ keratinocyte infiltration and CESC, and the localization relationship between PI3+S100A7+ keratinocytes and macrophages was verified by immunofluorescence staining. RESULTS: Various types of keratinocytes and fibroblasts were the two cell types with the most significant differences in percentage between the tumor tissue samples and paired adjacent non-cancerous tissue samples in the early stages of CESC. We found that PI3+S100A7+ keratinocytes were associated with early HPV-positive CESC, and PI3+S100A7+ keratinocytes were more abundant in tumors than in adjacent normal tissues in the TCGA-CESC dataset. Analysis of clinical information revealed that the infiltration of PI3+S100A7+ keratinocytes was notably higher in tumors with poor prognosis than in those with good prognosis. Additionally, multiplex immunofluorescence analysis showed a specific increase in PI3+S100A7+ expression within tumor tissues, with PI3+S100A7+ keratinocytes and CD163+ macrophages being spatially very close to each other. In the analysis of cell-cell interactions, macrophages exhibited strong crosstalk with PI3+S100A7+ proliferating keratinocytes in HPV-positive CESC tumors, mediated by tumor necrosis factor (TNF), CCL2, CXCL8, and IL10, highlighting the dynamic and tumor-specific enhancement of macrophage-keratinocyte interactions, which are associated with poor prognosis and immune modulation. Using CIBERSORTx, we discovered that patients with high infiltration of both PI3+S100A7+ proliferating keratinocytes and macrophages had the shortest overall survival. In the analysis of cell-cell interactions, PI3+S100A7+ proliferating keratinocytes and macrophages were found to be involved in highly active pathways that promote differentiation and structure formation, including cytokine receptor interactions, the Nuclear Factor kappa-light-chain-enhancer of activated B cells (NF-κB) signaling pathway, and TNF signaling pathway regulation. Further subtyping of fibroblast populations identified four subtypes. The C1 group, characterized by its predominance in tumor tissues, is a subtype enriched with cancer-associated fibroblasts (CAFs), whereas the C3 group is primarily enriched in adjacent non-cancerous tissues and consists of undifferentiated cells. Moreover, the distinct molecular and cellular differences between HPV16- and HPV66-associated tumors were demonstrated, emphasizing the unique tumor-promoting mechanisms and microenvironmental influences driven by each HPV subtype. CONCLUSIONS We discovered a heterogeneous population of keratinocytes between tumor and adjacent non-cancerous tissues caused by HPV infection and identified macrophages and specific CAFs that play a crucial role during the early stage in promoting the inflammatory response and remodeling the cancer-promoting TME. Our findings provide new insights into the transcriptional landscape of early-stage CESC to understand the mechanism of HPV-mediated malignant transformation in cervical cancer.
Humans ; Female ; Papillomavirus Infections/genetics* ; Uterine Cervical Neoplasms/genetics* ; Carcinoma, Squamous Cell/pathology* ; Keratinocytes/metabolism* ; Single-Cell Analysis/methods* ; Tumor Microenvironment/genetics*

Objective To investigate the value of multimodal model integrating digital mammography(MG)and MRI radiomics features for preoperative prediction of lymphovascular invasion(LVI)status in breast cancer.Methods The clinical and imaging data from 336 patients with pathologically confirmed invasive breast cancer were retrospectively analyzed and randomly divided into a training group(235 cases)and a test group(101 cases)according to the ratio of 7∶3.Feature dimensionality reduction was carried out by Pearson correlation analysis followed by least absolute shrinkage and selection operator(LASSO)regression.Radiomics models were constructed based on MG craniocaudal(CC),dynamic contrast enhancement(DCE),T2 WI,and integrated MRI sequences;a multimodal model was further developed by incorporating clinical high-risk factors.The predictive efficiency of each model was evaluated by plotting receiver operating characteristic(ROC)curve.Results The ROC curve analysis showed that the multimodal model performed the best predictive efficiency,with area under the curve(AUC)of 0.989 and 0.861,accuracy of 0.949 and 0.782,sensitivity of 0.923 and 0.828,and specificity of 0.962 and 0.764 in the training group and test group respectively.Conclusion The multimodal model,integrating MG and MRI radiomics features,show optimal performance and can be served as a preoperative prediction of LVI status in breast cancer.

Objective To investigate the value of multimodal model integrating digital mammography(MG)and MRI radiomics features for preoperative prediction of lymphovascular invasion(LVI)status in breast cancer.Methods The clinical and imaging data from 336 patients with pathologically confirmed invasive breast cancer were retrospectively analyzed and randomly divided into a training group(235 cases)and a test group(101 cases)according to the ratio of 7∶3.Feature dimensionality reduction was carried out by Pearson correlation analysis followed by least absolute shrinkage and selection operator(LASSO)regression.Radiomics models were constructed based on MG craniocaudal(CC),dynamic contrast enhancement(DCE),T2 WI,and integrated MRI sequences;a multimodal model was further developed by incorporating clinical high-risk factors.The predictive efficiency of each model was evaluated by plotting receiver operating characteristic(ROC)curve.Results The ROC curve analysis showed that the multimodal model performed the best predictive efficiency,with area under the curve(AUC)of 0.989 and 0.861,accuracy of 0.949 and 0.782,sensitivity of 0.923 and 0.828,and specificity of 0.962 and 0.764 in the training group and test group respectively.Conclusion The multimodal model,integrating MG and MRI radiomics features,show optimal performance and can be served as a preoperative prediction of LVI status in breast cancer.

Objective:This study aimed to construct a prediction model for diagnosis of proliferative lupus nephritis based on a machine learning algorithm. Additionally, a user-friendly platform was developed to propose a non-invasive method to assist the pathologic classification of lupus nephritis.Methods:A retrospective analysis was conducted on clinical and pathological data of lupus nephritis patients confirmed by renal biopsy at Zhengzhou University People′s Hospital from January 2017 to August 2023. The study population was randomly divided into training and testing sets in a 7∶3 ratio. Utilizing six machine learning algorithms, classification models were developed. The predictive performance of each model was assessed using metrics such as accuracy, sensitivity, specificity, and the area under the receiver operating characteristic curve (AUC). The optimal model, once identified, was deployed as a web-based calculator for convenient model application. SPSS 25.0 and R 4.2.2 were used to analyze the data.Results:The study included a total of 212 patients, with 138 cases with proliferative lupus nephritis and 74 cases with non-proliferative lupus nephritis. The AUC values for the six models, namely logistic regression, decision tree, random forest, support vector machine, extreme gradient boosting, and light gradient boosting machine, were 0.79, 0.62, 0.79, 0.88, 0.81, and 0.77, respectively; the accuracy rates were 82.54%, 65.08%, 74.60%, 85.71%, 69.84%, 71.43%, respectively. Among them, the support vector machine model demonstrated the optimal performance. This model had deployed as a web-based calculator. Based on feature importance scores, the top 10 influencing factors were identified, including anti URNP antibody, immunoglobulin G, serum globulin, estimated glomerular filtration rate, anti Smith antibody, BMI index, anti dsDNA antibody, uric acid, anti-Rib.p antibody, and gender.Conclusion:A prediction model based on machine learning algorithms was successfully established, and a web calculator was developed to offer a simple and non-invasive method for diagnosing proliferative lupus nephritis. This can assist clinicians in evaluating the risk-benefit ratio of kidney biopsy in patients with lupus nephritis.

Objective:To investigate the diagnostic performance of different thyroid imaging reporting and data systems (TI-RADS) in the diagnosis of medullary thyroid carcinoma (MTC) .Methods:A total of 160 thyroid nodules diagnosed as MTC by postoperative pathology from Aug. 2011 to Aug. 2022 at the First Affiliated Hospital of Zhengzhou University were included. Additionally, 160 papillary thyroid carcinomas (PTC) and 160 benign nodules were randomly selected as controls during the same period. Differences in gender, age, nodule diameter and various ultrasound features were observed. The nodules were classified according to American College of Radiology (ACR) TI-RADS, artificial intelligence (AI) TI-RADS and Chinese (C TI-RADS). Then, receiver operating characteristic curve (ROC) were plotted to calculate the diagnostic value. The Kendall concordance coefficient was used to evaluate the interobserver consistency of each TI-RADS system.Results:There was no statistically significant difference in gender among the three groups ( χ2=1.17, P=0.558). However, significant differences were observed in age and nodule diameter ( F=12.08,40.12, P<0.001 for both). The area under ROC (AUC) for diagnosing MTC and benign nodules using ACR, AI, and C-TIRADS were 0.762, 0.773, and 0.761, respectively, with no statistically significant differences ( Z=1.33, 0.01, 0.87, P=0.183, 0.994, 0.386). However, the sensitivity of C TI-RADS (87.5%) was lower than that of ACR and AI TI-RADS (both 95.0%) ( P=0.018). After combining the biopsy threshold, the false negative rate of C-TIRADS was lower than that of ACR (30.6% vs. 41.3%) ( P=0.048) and AI TIRADS (30.6% vs. 43.1%) ( P=0.020). The inter-observer diagnostic consistency of C-TIRADS was superior to ACR (0.884 vs. 0.819, P<0.001) and AI TIRADS (0.884 vs. 0.839) ( P<0.001) . Conclusions:AI and ACR TI-RADS have higher sensitivity in diagnosing MTC, while C TI-RADS has a lower puncture missed diagnosis rate. AI has similar diagnostic performance to ACR TI-RADS and can replace ACR TI-RADS.

Purpose To identify hemangioma(HE)and Kaposiform hemangioendothelioma(KHE)by constructing two ultrasonography-based radiomics models to evaluate the application value of two models in distinguishing HE from KHE,and to compare the diagnostic efficiency of two models.Materials and Methods A total of 90 lesions of subcutaneous HE or KHE confirmed clinically or pathologically from Henan Provincial People's Hospital from August 2020 to May 2022,were retrospectively analyzed.Imaging features were extracted by using Pyradiomics and screened out by the least absolute shrinkage and selection operator algorithm.Support vector machine and random forest were used to construct the radiomics models.Then the diagnostic efficacy of different models was compared.Results Based on the selected 10 radiomics features,the area under the curve,accuracy,sensitivity,specificity,positive and negative prediction the training group and validation group of the support vector machine model were 0.902(95%CI 0.887-0.917),92.1%,85.0%,92.3%,90.9%,93.5%and 0.827(95%CI 0.787-0.856),85.2%,70.0%,94.1%,90.9%,85.0%,respectively;and those in the training group and validation group of the random forest model were 0.960(95%CI 0.938-0.983),98.4%,96.4%,97.8%,98.1%,97.2%and 0.742(95%CI 0.699-0.785),77.8%,57.1%,82.3%,79.6%,62.5%,respectively.The area under the curve between two models in the training group and validation group was statistically significant(Z=-3.306,-2.009;P<0.05).Conclusion Ultrasonography-based radiomics can distinguish HE from KHE,support vector machine model shows more stable diagnostic performance in small sample data.

Objective: To investigate the current situation of using virtual simulation technology in undergraduate schools of stomatology in China, analyze the problems and put forward corresponding improvement suggestions. Methods: A questionnaire survey was conducted among 672 teachers and 3 849 students in undergraduate schools of stomatology in China. Results : 25.81% of all participants had took part in dental virtual simulation courses, and 37.80% of the participants from “Double First-Class” universities had participated in dental virtual simulation courses. 92.12% of the virtual simulation courses were established for undergraduates. "Traditional course + virtual simulation model demonstration" is the main teaching form of virtual simulation courses. Most of the participants were satisfied with the virtual simulation courses offered by their schools. At present, there are also some deficiencies in the virtual simulation courses, such as lack of teaching resources, insufficient interaction and simulation. Conclusion There is difference in the application of virtual simulation technology in undergraduate schools of stomatology in China. The virtual simulation technology is more widely used in "Double First-Class" universities than in ordinary universities. Undergraduates are the main teaching objects of virtual simulation courses. Stomatological schools in China should pay attention to the development and utilization of virtual simulation curriculum resources by cooperation, enrich the form of virtual simulation courses and strengthen the promotion and application of virtual simulation technology in stomatological education.

After decades of development,protein and peptide drugs have now grown into a major drug class in the marketplace.Target identification and validation are crucial for the discovery of protein and peptide drugs,and bioinformatics prediction of targets based on the characteristics of known target proteins will help improve the efficiency and success rate of target selection.However,owing to the developmental history in the pharmaceutical industry,previous systematic exploration of the target spaces has mainly focused on traditional small-molecule drugs,while studies related to protein and peptide drugs are lacking.Here,we systematically explore the target spaces in the human genome specifically for protein and peptide drugs.Compared with other proteins,both suc-cessful protein and peptide drug targets have many special characteristics,and are also significantly different from those of small-molecule drugs in many aspects.Based on these features,we develop separate effective genome-wide target prediction models for protein and peptide drugs.Finally,a user-friendly web server,Predictor Of Protein and Peptide drugs'therapeutic Targets(POPPIT)(http://poppit.ncpsb.org.cn/),is established,which provides not only target prediction specifically for protein and peptide drugs but also abundant annotations for predicted targets.

Objective To detect mutations in the ARAD1 gene in a pedigree with dyschromatosis symmetrica hereditaria (DSH).Methods Genomic DNA was extracted from the peripheral blood of 8 family members (including 5 patients with DSH and 3 unaffected members) in the pedigree with DSH,as well as 100 unrelated healthy controls.All the 15 exon sequences of the ADAR1 gene were amplified by polymerase chain reaction (PCR)followed by direct sequencing.Then,mutations were detected in comparison with the standard sequence of the ADAR1 gene in Genebank.Results A nonsense mutation C.1420C ＞ T (p.Arg474X) was identified at position 1 420 in exon 2 of the ADAR1 gene in the 5 patients with DSH,but not in the 3 unaffected members or 100 unrelated healthy controls.Conclusion The nonsense mutation C.1420C ＞ T in the ADAR1 gene is the causative mutation in the pedigree with DSH.