Application of machine learning algorithms to identify people with low bone density.

Background: Osteoporosis is becoming more common worldwide, imposing a substantial burden on individuals and society. The onset of osteoporosis is subtle, early detection is challenging, and population-wide screening is infeasible. Thus, there is a need to develop a method to identify those at high risk for osteoporosis. Objective: This study aimed to develop a machine learning algorithm to effectively identify people with low bone density, using readily available demographic and blood biochemical data. Methods: Using NHANES 2017-2020 data, participants over 50 years old with complete femoral neck BMD data were selected. This cohort was randomly divided into training (70%) and test (30%) sets. Lasso regression selected variables for inclusion in six machine learning models built on the training data: logistic regression (LR), support vector machine (SVM), gradient boosting machine (GBM), naive Bayes (NB), artificial neural network (ANN) and random forest (RF). NHANES data from the 2013-2014 cycle was used as an external validation set input into the models to verify their generalizability. Model discrimination was assessed via AUC, accuracy, sensitivity, specificity, precision and F1 score. Calibration curves evaluated goodness-of-fit. Decision curves determined clinical utility. The SHAP framework analyzed variable importance. Results: A total of 3,545 participants were included in the internal validation set of this study, of whom 1870 had normal bone density and 1,675 had low bone density Lasso regression selected 19 variables. In the test set, AUC was 0.785 (LR), 0.780 (SVM), 0.775 (GBM), 0.729 (NB), 0.771 (ANN), and 0.768 (RF). The LR model has the best discrimination and a better calibration curve fit, the best clinical net benefit for the decision curve, and it also reflects good predictive power in the external validation dataset The top variables in the LR model were: age, BMI, gender, creatine phosphokinase, total cholesterol and alkaline phosphatase. Conclusion: The machine learning model demonstrated effective classification of low BMD using blood biomarkers. This could aid clinical decision making for osteoporosis prevention and management.

Health status of the population in Naqu, Tibet and its latent class analysis: a cross-sectional survey.

Background: Through a survey and analysis of the population's present state of health, it is possible to give data support for improving the health status of inhabitants in Naqu, Tibet. Additionally, it is possible to provide specific recommendations for the development of medical and healthcare facilities in Tibet. Methods: The health scores of the participants were based on their responses to the four main sections of the questionnaire: dietary habits, living habits, health knowledge, and clinical disease history, and the variability of health status among groups with different characteristics was analyzed based on the scores. The four major sections were used to create classes of participants using latent class analysis (LCA). Using logistic regression, the factors influencing the classification of latent classes of health status were investigated. Results: A total of 995 residents from 10 counties in Naqu were selected as the study subjects. And their demographic characteristics were described. The mean health score of residents after standardization was 81.59 ± 4.68. With the exception of gender, health scores differed between groups by age, education level, different occupations, marital status, and monthly income. The health status in Naqu, Tibet, was divided into two groups (entropy = 0.29, BLRT = 0.001, LMRT = 0.001) defined as the "good health group" and the "general health group." A monthly income of more than ¥5000 adverse to good health in Naqu, Tibet. Discussion: Single, well-educated young adults in Naqu, Tibet, have outstanding health. The vast majority of people in Tibet's Naqu region were in good health. Furthermore, the population's latent health status was divided into two classes, each with good dietary and living habits choices, low health knowledge, and a history of several clinical diseases. Univariate and multivariate logistic regression analysis showed that monthly income more than ¥5000 was an independent risk factor for poor health status.

Integrating Single-Cell RNA-Seq and Bulk RNA-Seq Data to Explore the Key Role of Fatty Acid Metabolism in Breast Cancer.

Cancer immune escape is associated with the metabolic reprogramming of the various infiltrating cells in the tumor microenvironment (TME), and combining metabolic targets with immunotherapy shows great promise for improving clinical outcomes. Among all metabolic processes, lipid metabolism, especially fatty acid metabolism (FAM), plays a major role in cancer cell survival, migration, and proliferation. However, the mechanisms and functions of FAM in the tumor immune microenvironment remain poorly understood. We screened 309 fatty acid metabolism-related genes (FMGs) for differential expression, identifying 121 differentially expressed genes. Univariate Cox regression models in The Cancer Genome Atlas (TCGA) database were then utilized to identify the 15 FMGs associated with overall survival. We systematically evaluated the correlation between FMGs' modification patterns and the TME, prognosis, and immunotherapy. The FMGsScore was constructed to quantify the FMG modification patterns using principal component analysis. Three clusters based on FMGs were demonstrated in breast cancer, with three patterns of distinct immune cell infiltration and biological behavior. An FMGsScore signature was constructed to reveal that patients with a low FMGsScore had higher immune checkpoint expression, higher immune checkpoint inhibitor (ICI) scores, increased immune microenvironment infiltration, better survival advantage, and were more sensitive to immunotherapy than those with a high FMGsScore. Finally, the expression and function of the signature key gene NDUFAB1 were examined by in vitro experiments. This study significantly demonstrates the substantial impact of FMGs on the immune microenvironment of breast cancer, and that FMGsScores can be used to guide the prediction of immunotherapy efficacy in breast cancer patients. In vitro experiments, knockdown of the NDUFAB1 gene resulted in reduced proliferation and migration of MCF-7 and MDA-MB-231 cell lines.

Single-cell sequencing and bulk RNA data reveal the tumor microenvironment infiltration characteristics of disulfidptosis related genes in breast cancer.

BACKGROUND: Immunotherapy, represented by immune checkpoint inhibitors, has made significant progress in the treatment of cancer. Numerous studies have demonstrated that antitumor therapies targeting cell death exhibit synergistic effects with immunotherapy. Disulfidptosis is a recently discovered form of cell death, and its potential influence on immunotherapy, similar to other regulated cell death processes, requires further investigation. The prognostic value of disulfidptosis in breast cancer and its role in the immune microenvironment has not been investigated. METHODS: High dimensional weighted gene coexpression network analysis (hdWGCNA) and Weighted co-expression network analysis (WGCNA) methods were employed to integrate breast cancer single-cell sequencing data and bulk RNA data. These analyses aimed to identify genes associated with disulfidptosis in breast cancer. Risk assessment signature was constructed using Univariate Cox and least absolute shrinkage and selection operator (LASSO) analyses. RESULTS: In this study, we constructed a risk signature by disulfidptosis-related genes to predict overall survival and immunotherapy response in BRCA patients. The risk signature demonstrated robust prognostic power and accurately predicted survival compared to traditional clinicopathological features. It also effectively predicted the response to immunotherapy in patients with breast cancer. Through cell communication analysis in additional single-cell sequencing data, we identified TNFRSF14 as a key regulatory gene. Combining TNFRSF14 targeting and immune checkpoint inhibition to induce disulfidptosis in tumor cells could potentially suppress tumor proliferation and enhance survival in patients with BRCA.

Risk factor analysis and nomogram establishment and verification of brain astrocytoma patients based on SEER database.

Astrocytoma is a common brain tumor that can occur in any part of the central nervous system. This tumor is extremely harmful to patients, and there are no clear studies on the risk factors for astrocytoma of the brain. This study was conducted based on the SEER database to determine the risk factors affecting the survival of patients with astrocytoma of the brain. Patients diagnosed with brain astrocytoma in the SEER database from 2004 to 2015 were screened by inclusion exclusion criteria. Final screened brain astrocytoma patients were classified into low grade and high grade according to WHO classification. The risk factors affecting the survival of patients with low-grade and high-grade brain astrocytoma were analyzed by univariate Kaplan-Meier curves and log-rank tests, individually. Secondly, the data were randomly divided into training set and validation set according to the ratio of 7:3, and the training set data were analyzed by univariate and multivariate Cox regression, and the risk factors affecting the survival of patients were screened and nomogram was established to predict the survival rates of patients at 3 years and 5 years. The area under the ROC curve (AUC value), C-index, and Calibration curve are used to evaluate the sensitivity and calibration of the model. Univariate Kaplan-Meier survival curve and log-rank test showed that the risk factors affecting the prognosis of patients with low-grade astrocytoma included Age, Primary site, Tumor histological type, Grade, Tumor size, Extension, Surgery, Radiation, Chemotherapy and Tumor number; risk factors affecting the prognosis of patients with high-grade astrocytoma include Age, Primary site, Tumor histological type, Tumor size, Extension, Laterality, Surgery, Radiation, Chemotherapy and Tumor number. Through Cox regression, independent risk factors of patients with two grades were screened separately, and nomograms of risk factors for low-grade and high-grade astrocytoma were successfully established to predict the survival rate of patients at 3 and 5 years. The AUC values of low-grade astrocytoma training set patients were 0.829 and 0.801, and the C-index was 0.818 (95% CI 0.779, 0.857). The AUC values of patients in the validation set were 0.902, 0.829, and the C-index was 0.774 (95% CI 0.758, 0.790), respectively. The AUC values of high-grade astrocytoma training set patients were 0.814 and 0.806, the C-index was 0.774 (95% CI 0.758, 0.790), the AUC values of patients in the validation set were 0.802 and 0.823, and the C-index was 0.766 (95% CI 0.752, 0.780), respectively, and the calibration curves of the two levels of training set and validation set were well fitted. This study used data from the SEER database to identify risk factors affecting the survival prognosis of patients with brain astrocytoma, which can provide some guidance for clinicians.

Association of dietary pattern and Tibetan featured foods with high-altitude polycythemia in Naqu, Tibet: A 1:2 individual-matched case-control study.

This study focused on the association of dietary patterns and Tibetan featured foods with high-altitude polycythemia (HAPC) in Naqu, Tibet, to explore the risk factors of HAPC in Naqu, Tibet, to raise awareness of the disease among the population and provide evidence for the development of prevention and treatment interventions. A 1:2 individual-matched case-control study design was used to select residents of three villages in the Naqu region of Tibet as the study population. During the health examination and questionnaire survey conducted from December 2020 to December 2021, a sample of 1,171 cases was collected. And after inclusion and exclusion criteria and energy intake correction, 100 patients diagnosed with HAPC using the "Qinghai criteria" were identified as the case group, while 1,059 patients without HAPC or HAPC -related diseases were identified as the control group. Individuals were matched by a 1:2 propensity score matching according to gender, age, body mass index (BMI), length of residence, working altitude, smoking status, and alcohol status. Dietary patterns were determined by a principal component analysis, and the scores of study subjects for each dietary pattern were calculated. The effect of dietary pattern scores and mean daily intake (g/day) of foods in the Tibetan specialty diet on the prevalence of HAPC was analyzed using conditional logistic regression. After propensity score matching, we found three main dietary patterns among residents in Naqu through principal component analysis, which were a "high protein pattern," "snack food pattern," and "vegetarian food pattern." All three dietary patterns showed a high linear association with HAPC (p < 0.05) and were risk factors for HAPC. In the analysis of the relationship between Tibetan featured foods and the prevalence of HAPC, the results of the multifactorial analysis following adjustment for other featured foods showed that there was a positive correlation between the average daily intake of tsampa and the presence of HAPC, which was a risk factor. Additionally, there was an inverse correlation between the average daily intake of ghee tea and the presence of HAPC, which was a protective factor.