Multimorbidity measures associated with cognitive function among community-dwelling older Chinese adults.

INTRODUCTION: Older adults with multimorbidity are at high risk of cognitive impairment development. There is a lack of research on the associations between different multimorbidity measures and cognitive function among older Chinese adults living in the community. METHODS: We used the Chinese Longitudinal Healthy Longevity Survey from 2002 to 2018 and included data on dementia-free participants aged ≥65 years. Multimorbidity measures included condition counts, multimorbidity patterns, and trajectories. The association of multimorbidity measures with cognitive function was examined by generalized estimating equation and linear and logistic regression models. RESULTS: Among 14,093 participants at baseline, 43.2% had multimorbidity. Multimorbidity patterns were grouped into cancer-inflammatory, cardiometabolic, and sensory patterns. Multimorbidity trajectories were classified as "onset-condition," "newly developing," and "severe condition." The Mini-Mental State Examination scores were significantly lower for participants with more chronic conditions, with cancer-inflammatory/cardiometabolic/sensory patterns, and with developing multimorbidity trajectories. DISCUSSION: Condition counts, sensory pattern, cardiometabolic pattern, cancer-inflammatory pattern, and multimorbidity developmental trajectories were prospectively associated with cognitive function. HIGHLIGHTS: Elderly individuals with a higher number of chronic conditions were associated with lower MMSE scores in the Chinese Longitudinal Healthy Longevity Survey data. MMSE scores were significantly lower for participants with specific multimorbidity patterns. Individuals with developing trajectories of multimorbidity were associated with lower MMSE scores and a higher risk of mild cognitive impairment.

Deep learning model to discriminate diverse infection types based on pairwise analysis of host gene expression.

Accurate detection of pathogens, particularly distinguishing between Gram-positive and Gram-negative bacteria, could improve disease treatment. Host gene expression can capture the immune system's response to infections caused by various pathogens. Here, we present a deep neural network model, bvnGPS2, which incorporates the attention mechanism based on a large-scale integrated host transcriptome dataset to precisely identify Gram-positive and Gram-negative bacterial infections as well as viral infections. We performed analysis of 4,949 blood samples across 40 cohorts from 10 countries using our previously designed omics data integration method, iPAGE, to select discriminant gene pairs and train the bvnGPS2. The performance of the model was evaluated on six independent cohorts comprising 374 samples. Overall, our deep neural network model shows robust capability to accurately identify specific infections, paving the way for precise medicine strategies in infection treatment and potentially also for identifying subtypes of other diseases.

Integrative machine learning and neural networks for identifying PANoptosis-related lncRNA molecular subtypes and constructing a predictive model for head and neck squamous cell carcinoma.

PURPOSE: PANoptosis is considered a novel type of cell death that plays important roles in tumor progression. In this study, we applied machine learning algorithms to explore the relationships between PANoptosis-related lncRNAs (PRLs) and head and neck squamous cell carcinoma (HNSCC) and established a neural network model for prognostic prediction. METHODS: Information about the HNSCC cohort was downloaded from the TCGA database, and the differentially expressed prognostic PRLs between tumor and normal samples were assessed in patients with different tumor subtypes via nonnegative matrix factorization (NMF) analysis. Subsequently, five kinds of machine-learning algorithms were used to select the core PRLs across the subtypes, and the interactive features were pooled into a neural network model to establish a PRL-related risk score (PLRS) system. Survival differences were compared via Kaplan‒Meier analysis, and the predictive effects were assessed with the areas under the ROCs. Moreover, functional enrichment analysis, immune infiltration, tumor mutation burden (TMB) and clinical therapeutic response were also conducted to further evaluate the novel predictive model. RESULTS: A total of 347 PRLs were identified, 225 of which were differentially expressed between tumor and normal samples. Patients were divided into two clusters via NMF analysis, in which cluster 1 had a better prognosis and more immune cells and functional infiltrates. With the application of five machine learning algorithms, we selected 13 interactive PRLs to construct the predictive model. The AUCs for the ROCs in the entire set were 0.735, 0.740 and 0.723, respectively. Patients in the low-PLRS group exhibited a better prognosis, greater immune cell enrichment, greater immune function activation, lower TMB and greater sensitivity to immunotherapy. CONCLUSION: In this study, we established a novel neural network prognostic model to predict survival and identify tumor subtypes in HNSCC patients. This novel assessment system is useful for prediction, providing ideas for clinical treatment.

Deregulation of immune response contributing to fulminant hepatitis in HEV infected pregnant women.

Hepatitis E virus (HEV) infection in pregnant women is associated with a wide spectrum of adverse consequences for both mother and fetus. The high mortality in this population appears to be associated with hormonal changes and consequent immunological changes. This study conducted an analysis of immune responses in pregnant women infected with HEV manifesting varying severity. Data mining analysis of the GSE79197 was utilized to examine differentially biological functions in pregnant women with HEV infection (P-HEV) versus without HEV infection (P-nHEV), P-HEV progressing to ALF (P-ALF) versus P-HEV, and P-HEV versus non-pregnant women with HEV infection (nP-HEV). We found cellular response to interleukin and immune response-regulating signalings were activated in P-HEV compared with P-nHEV. However, there was a significant decrease of immune responses, such as T cell activation, leukocyte cell-cell adhesion, regulation of lymphocyte activation, and immune response-regulating signaling pathway in P-ALF patient than P-HEV patient. Compared with nP-HEV, MHC protein complex binding function was inhibited in P-HEV. Further microRNA enrichment analysis showed that MAPK and T cell receptor signaling pathways were inhibited in P-HEV compared with nP-HEV. In summary, immune responses were activated during HEV infection while being suppressed when developing ALF during pregnancy, heightening the importance of immune mediation in the pathogenesis of severe outcome in HEV infected pregnant women.

Commonly used software tools produce conflicting and overly-optimistic AUPRC values.

The precision-recall curve (PRC) and the area under the precision-recall curve (AUPRC) are useful for quantifying classification performance. They are commonly used in situations with imbalanced classes, such as cancer diagnosis and cell type annotation. We evaluate 10 popular tools for plotting PRC and computing AUPRC, which were collectively used in more than 3000 published studies. We find the AUPRC values computed by the tools rank classifiers differently and some tools produce overly-optimistic results.

Targeting adipocyte ESRRA promotes osteogenesis and vascular formation in adipocyte-rich bone marrow.

Excessive bone marrow adipocytes (BMAds) accumulation often occurs under diverse pathophysiological conditions associated with bone deterioration. Estrogen-related receptor α (ESRRA) is a key regulator responding to metabolic stress. Here, we show that adipocyte-specific ESRRA deficiency preserves osteogenesis and vascular formation in adipocyte-rich bone marrow upon estrogen deficiency or obesity. Mechanistically, adipocyte ESRRA interferes with E2/ESR1 signaling resulting in transcriptional repression of secreted phosphoprotein 1 (Spp1); yet positively modulates leptin expression by binding to its promoter. ESRRA abrogation results in enhanced SPP1 and decreased leptin secretion from both visceral adipocytes and BMAds, concertedly dictating bone marrow stromal stem cell fate commitment and restoring type H vessel formation, constituting a feed-forward loop for bone formation. Pharmacological inhibition of ESRRA protects obese mice against bone loss and high marrow adiposity. Thus, our findings highlight a therapeutic approach via targeting adipocyte ESRRA to preserve bone formation especially in detrimental adipocyte-rich bone milieu.

Analysis of differential metabolites in serum metabolomics of patients with aortic dissection.

BACKGROUND: Pathogenesis and diagnostic biomarkers of aortic dissection (AD) can be categorized through the analysis of differential metabolites in serum. Analysis of differential metabolites in serum provides new methods for exploring the early diagnosis and treatment of aortic dissection. OBJECTIVES: This study examined affected metabolic pathways to assess the diagnostic value of metabolomics biomarkers in clients with AD. METHOD: The serum from 30 patients with AD and 30 healthy people was collected. The most diagnostic metabolite markers were determined using metabolomic analysis and related metabolic pathways were explored. RESULTS: In total, 71 differential metabolites were identified. The altered metabolic pathways included reduced phospholipid catabolism and four different metabolites considered of most diagnostic value including N2-gamma-glutamylglutamine, PC(phocholines) (20:4(5Z,8Z,11Z,14Z)/15:0), propionyl carnitine, and taurine. These four predictive metabolic biomarkers accurately classified AD patient and healthy control (HC) samples with an area under the curve (AUC) of 0.9875. Based on the value of the four different metabolites, a formula was created to calculate the risk of aortic dissection. Risk score = (N2-gamma-glutamylglutamine × -0.684) + (PC (20:4(5Z,8Z,11Z,14Z)/15:0) × 0.427) + (propionyl carnitine × 0.523) + (taurine × -1.242). An additional metabolic pathways model related to aortic dissection was explored. CONCLUSION: Metabolomics can assist in investigating the metabolic disorders associated with AD and facilitate a more in-depth search for potential metabolic biomarkers.

Intra-sample reversed pairs based on differentially ranked genes reveal biosignature for ovarian cancer.

Ovarian cancer, a major gynecological malignancy, often remains undetected until advanced stages, necessitating more effective early screening methods. Existing biomarker based on differential genes often suffers from variations in clinical practice. To overcome the limitations of absolute gene expression values including batch effects and biological heterogeneity, we introduced a pairwise biosignature leveraging intra-sample differentially ranked genes (DRGs) and machine learning for ovarian cancer detection across diverse cohorts. We analyzed ten cohorts comprising 872 samples with 796 ovarian cancer and 76 normal. Our method, DRGpair, involves three stages: intra-sample ranking differential analysis, reversed gene pair analysis, and iterative LASSO regression. We identified four DRG pairs, demonstrating superior diagnostic performance compared to current state-of-the-art biomarkers and differentially expressed genes in seven independent cohorts. This rank-based approach not only reduced computational complexity but also enhanced the specificity and effectiveness of biomarkers, revealing DRGs as promising candidates for ovarian cancer detection and offering a scalable model adaptable to varying cohort characteristics.

DeepLocRNA: an interpretable deep learning model for predicting RNA subcellular localization with domain-specific transfer-learning.

MOTIVATION: Accurate prediction of RNA subcellular localization plays an important role in understanding cellular processes and functions. Although post-transcriptional processes are governed by trans-acting RNA binding proteins (RBPs) through interaction with cis-regulatory RNA motifs, current methods do not incorporate RBP-binding information. RESULTS: In this article, we propose DeepLocRNA, an interpretable deep-learning model that leverages a pre-trained multi-task RBP-binding prediction model to predict the subcellular localization of RNA molecules via fine-tuning. We constructed DeepLocRNA using a comprehensive dataset with variant RNA types and evaluated it on the held-out dataset. Our model achieved state-of-the-art performance in predicting RNA subcellular localization in mRNA and miRNA. It has also demonstrated great generalization capabilities, performing well on both human and mouse RNA. Additionally, a motif analysis was performed to enhance the interpretability of the model, highlighting signal factors that contributed to the predictions. The proposed model provides general and powerful prediction abilities for different RNA types and species, offering valuable insights into the localization patterns of RNA molecules and contributing to our understanding of cellular processes at the molecular level. A user-friendly web server is available at: https://biolib.com/KU/DeepLocRNA/.

Commonly used software tools produce conflicting and overly-optimistic AUPRC values.

The precision-recall curve (PRC) and the area under it (AUPRC) are useful for quantifying classification performance. They are commonly used in situations with imbalanced classes, such as cancer diagnosis and cell type annotation. We evaluated 10 popular tools for plotting PRC and computing AUPRC, which were collectively used in >3,000 published studies. We found the AUPRC values computed by the tools rank classifiers differently and some tools produce overly-optimistic results.

Diagnostic Prediction of portal vein thrombosis in chronic cirrhosis patients using data-driven precision medicine model.

BACKGROUND: Portal vein thrombosis (PVT) is a significant issue in cirrhotic patients, necessitating early detection. This study aims to develop a data-driven predictive model for PVT diagnosis in chronic hepatitis liver cirrhosis patients. METHODS: We employed data from a total of 816 chronic cirrhosis patients with PVT, divided into the Lanzhou cohort (n = 468) for training and the Jilin cohort (n = 348) for validation. This dataset encompassed a wide range of variables, including general characteristics, blood parameters, ultrasonography findings and cirrhosis grading. To build our predictive model, we employed a sophisticated stacking approach, which included Support Vector Machine (SVM), Naïve Bayes and Quadratic Discriminant Analysis (QDA). RESULTS: In the Lanzhou cohort, SVM and Naïve Bayes classifiers effectively classified PVT cases from non-PVT cases, among the top features of which seven were shared: Portal Velocity (PV), Prothrombin Time (PT), Portal Vein Diameter (PVD), Prothrombin Time Activity (PTA), Activated Partial Thromboplastin Time (APTT), age and Child-Pugh score (CPS). The QDA model, trained based on the seven shared features on the Lanzhou cohort and validated on the Jilin cohort, demonstrated significant differentiation between PVT and non-PVT cases (AUROC = 0.73 and AUROC = 0.86, respectively). Subsequently, comparative analysis showed that our QDA model outperformed several other machine learning methods. CONCLUSION: Our study presents a comprehensive data-driven model for PVT diagnosis in cirrhotic patients, enhancing clinical decision-making. The SVM-Naïve Bayes-QDA model offers a precise approach to managing PVT in this population.

MrGPS: an m6A-related gene pair signature to predict the prognosis and immunological impact of glioma patients.

N6-methyladenosine (m6A) RNA methylation is the predominant epigenetic modification for mRNAs that regulates various cancer-related pathways. However, the prognostic significance of m6A modification regulators remains unclear in glioma. By integrating the TCGA lower-grade glioma (LGG) and glioblastoma multiforme (GBM) gene expression data, we demonstrated that both the m6A regulators and m6A-target genes were associated with glioma prognosis and activated various cancer-related pathways. Then, we paired m6A regulators and their target genes as m6A-related gene pairs (MGPs) using the iPAGE algorithm, among which 122 MGPs were significantly reversed in expression between LGG and GBM. Subsequently, we employed LASSO Cox regression analysis to construct an MGP signature (MrGPS) to evaluate glioma prognosis. MrGPS was independently validated in CGGA and GEO glioma cohorts with high accuracy in predicting overall survival. The average area under the receiver operating characteristic curve (AUC) at 1-, 3- and 5-year intervals were 0.752, 0.853 and 0.831, respectively. Combining clinical factors of age and radiotherapy, the AUC of MrGPS was much improved to around 0.90. Furthermore, CIBERSORT and TIDE algorithms revealed that MrGPS is indicative for the immune infiltration level and the response to immune checkpoint inhibitor therapy in glioma patients. In conclusion, our study demonstrated that m6A methylation is a prognostic factor for glioma and the developed prognostic model MrGPS holds potential as a valuable tool for enhancing patient management and facilitating accurate prognosis assessment in cases of glioma.

Accurate identification of structural variations from cancer samples.

Structural variations (SVs) are commonly found in cancer genomes. They can cause gene amplification, deletion and fusion, among other functional consequences. With an average read length of hundreds of kilobases, nano-channel-based optical DNA mapping is powerful in detecting large SVs. However, existing SV calling methods are not tailored for cancer samples, which have special properties such as mixed cell types and sub-clones. Here we propose the Cancer Optical Mapping for detecting Structural Variations (COMSV) method that is specifically designed for cancer samples. It shows high sensitivity and specificity in benchmark comparisons. Applying to cancer cell lines and patient samples, COMSV identifies hundreds of novel SVs per sample.