Dysfunction of protein trafficking has been intensively associated with neurological diseases, including neurodegeneration, but whether and how protein transport contributes to oligodendrocyte (OL) maturation and myelin repair in white matter injury remains unclear. ER-to-Golgi trafficking of newly synthesized proteins is mediated by coat protein complex II (COPII). Here, we demonstrate that the COPII component Sec13 was essential for OL differentiation and postnatal myelination. Ablation of Sec13 in the OL lineage prevented OPC differentiation and inhibited myelination and remyelination after demyelinating injury in the central nervous system (CNS), while improving protein trafficking by tauroursodeoxycholic acid (TUDCA) or ectopic expression of COPII components accelerated myelination. COPII components were upregulated in OL lineage cells after demyelinating injury. Loss of Sec13 altered the secretome of OLs and inhibited the secretion of pleiotrophin (PTN), which was found to function as an autocrine factor to promote OL differentiation and myelin repair. These data suggest that Sec13-dependent protein transport is essential for OL differentiation and that Sec13-mediated PTN autocrine signaling is required for proper myelination and remyelination.

In intensive care units (ICUs), mortality prediction is performed by combining information from these two sources of ICU patients by monitoring patient health. Respectively, time series data generated from each patient admission to the ICU and clinical records consisting of physician diagnostic summaries. However, existing mortality prediction studies mainly cascade the multimodal features of time series data and clinical records for prediction, ignoring thecross-modal correlation between the underlying features in different modal data. To address theseissues, we propose a multimodal fusion model for mortality prediction that jointly models patients’ time-series data as well as clinical records. We apply a fine-tuned Bert model (Bio-Bert) to the patient’s clinical record to generate a holistic embedding of the text part, which is then combined with the output of an LSTM model encoding the patient’s time-series data to extract valid features. The global contextual information of each modal data is extracted using the improved fusion module to capture the correlation between different modal data. Furthermore, the improved fusion module can be easily added to the fusion features of any unimodal network and utilize existing pre-trained unimodal model weights. We use a real dataset containing 18904 ICU patients to train and evaluate our model, and the research results show that the representations obtained by themodel can achieve better prediction accuracy compared to the baseline.

High-frequency oscillations (HFOs), observed within 80–500 Hz of magnetoencephalography (MEG) data, are putative biomarkers to localize epileptogenic zones that are critical for the success of surgical epilepsy treatment. It is crucial to accurately detect HFOs for improving the surgical outcome of patients with epilepsy. However, in clinical practices, detecting HFOs in MEG signals mainly depends on visual inspection by clinicians, which is very time-consuming, labor-intensive, subjective, and error-prone. To accurately and automatically detect HFOs, machine learning approaches have been developed and have demonstrated the promising results of automated HFO detection. More recently, the transformer-based model has attracted wide attention and achieved state-of-the-art performance on many machine learning tasks. In this paper, we are investigating the suitability of transformer-based models on the detection of HFOs. Specifically, we propose a transformer-based HFO detection framework for biomedical MEG one-dimensional signal data. For signal classification, we develop a transformer-based HFO (TransHFO) classification model. Then, we investigate the relationship between depth of deep learning models and classification performance. The experimental results show that the proposed framework outperforms the state-of-the-art HFO classifiers, increasing classification accuracy by 7%. Furthermore, we find that shallow TransHFO (<10 layers) outperforms deep TransHFO models (≥10 layers) on most data augmented factors.

Clinical diagnosis of epilepsy significantly relies on identifying interictal epileptiform discharge (IED) in electroencephalogram (EEG). IED is generally interpreted manually, and the related process is very time-consuming. Meanwhile, the process is expert-biased, which can easily lead to missed diagnosis and misdiagnosis. In recent years, with the development of deep learning, related algorithms have been used in automatic EEG analysis, but there are still few attempts in IED detection. This study uses the currently most popular convolutional neural network (CNN) framework for EEG analysis for automatic IED detection. The research topic is transferred into a 4-labels classification problem. The algorithm is validated on the long-term EEG of 11 pediatric patients with epilepsy. The computational results confirm that the CNN-based model can obtain high classification accuracy, up to 87%. The study may provide a reference for the future application of deep learning in automatic IED detection.

Background: With the aging population, the number of individuals with dementia in China is increasing rapidly. This community-based study aimed to investigate the prevalence and risk factors for dementia and mild cognitive impairment (MCI) among older adults in China. Methods: In this study, 20,070 individuals aged ≥ 65 were recruited between January 1, 2022, and February 1, 2023, from ten communities in Xiamen City, China. We collected data on age, sex, level of education, and medical history, as well as global cognition and functional status. The prevalence of dementia and MCI was examined, and the risk factors for different groups were assessed. Results: The overall prevalence of dementia and MCI was approximately 5.4% (95% confidence interval [CI], 5.1–5.7) and 7.7% (95% CI, 7.4–8.1), respectively. The results also indicated that dementia and MCI share similar risk factors, including older age, female sex, hypertension, and diabetes mellitus. Compared with individuals with no formal education, those with > 6 years of education had an odds ratio for MCI of 1.83 (95% CI, 1.49–2.25). We also found that only 5.5% of the positive participants chose to be referred to the hospital for further diagnosis and treatment during follow-up visits. Conclusions: This study estimated the prevalence and risk factors for dementia and MCI among individuals aged ≥ 65 years in Southeast China. These findings are crucial for preventing and managing dementia and MCI in China.

Background: Cuproptosis is increasingly recognized as an essential factor in the pathological process of Alzheimer’s disease (AD). However, the specific role of cuproptosis-related genes in AD remains poorly understood. Methods: Our first step was to obtain gene expression data from the GEO database and identify differentially expressed cuproptosis-associated genes (DECAGs) in AD. GO, KEGG, and GSEA analyses were then conducted on these genes. Subsequently, we attempted to classify AD patients by unsupervised clustering. Then, four machine-learning models were used to screen hub-genes from the DECAGs. We also explored the immune features of these genes and predicted target drugs. Molecular docking analysis was then performed on the predicted drugs and their corresponding hub-gene related proteins. Candidate markers were then validated by single-cell analysis and intracellular communication was investigated in a GEO scRNA-seq dataset. Lastly, we examined the expression levels of the hub-genes in peripheral blood cells using real-time quantitative PCR. Results: 19 DECAGs were found in AD and the key biological processes and molecular functions associated with AD were further determined. Two subtypes of peripheral blood cells showed significant alternations in AD: Cluster1 and Cluster2. Five hub-genes including FDX1, GLS, PDK1, MAP2K1, and SOD1 were then screened out from the machine-learning study. All of the five hub-genes were significantly correlated with various immunocytes. We discovered compounds targeting hub-gene related proteins and forecasted multiple strong hydrogen bonding interactions between the picked predicted drugs and the target proteins by molecular docking analysis. Subsequently, in the single-cell analysis of AD peripheral blood, all hub-genes except SOD1 were found to be up-regulated in B cells, NK cells, and CD4+ T cells, possibly acting on the MIF pathway. Finally, we discovered that the levels of PDK1 expression in AD patients were remarkably upregulated, while FDX1 and GLS were significantly decreased using qPCR. Conclusion: This study examined changes in intercellular communication between immune cells in the peripheral blood and identified five novel feature genes associated with cuproptosis in AD patients. These results facilitated a deeper understanding of the molecular mechanisms of AD and suggested novel therapeutic targets.