t-Distributed Stochastic Neighbor Embedding (t-SNE): A tool for eco-physiological transcriptomic analysis

An antibody developability triaging pipeline exploiting protein language models

Therapeutic monoclonal antibodies (mAbs) are a successful class of biologic drugs that are frequently selected from phage display libraries and transgenic mice that produce fully human antibodies. However, binding affinity to the correct epitope is necessary, but not sufficient, for a mAb to have therapeutic potential. Sequence and structural features affect the developability of an antibody, which influences its ability to be produced at scale and enter trials, or can cause late-stage failures. Using data on paired human antibody sequences, we introduce a pipeline using a machine learning approach that exploits protein language models to identify antibodies which cluster with antibodies that have entered the clinic and are therefore expected to have developability features similar to clinically acceptable antibodies, and triage out those without these features. We propose this pipeline as a useful tool in candidate selection from large libraries, reducing the cost of exploration of the antibody space, and pursuing new therapeutics.

Screening of estrogen receptor activity of per- and polyfluoroalkyl substances based on deep learning and in vivo assessment

Over the past decades, exposure to per- and polyfluoroalkyl substances (PFAS), a group of synthetic chemicals notorious for their environmental persistence, has been shown to pose increased health risks. Despite that some PFAS were reported to have endocrine-disrupting toxicity in previous studies, accurate prediction models based on deep learning and the underlying structural characteristics related to the effect of molecular fluorination remain limited. To address these issues, we proposed a stacking deep learning architecture, GXDNet, that integrates molecular descriptors and molecular graphs to predict the estrogen receptor α (ERα) activities of compounds, enhancing the generalization ability compared to previous models. Subsequently, we screened the ERα activity of 10,067 PFAS molecules using the GXDNet model and identified potential ERα binders. The representative PFAS molecules with the top docking scores showed that the introduction of fluorinated alkane chains significantly increased the binding affinities of parent molecules with ERα, suggesting that the combination of phenol structural fragments and fluorinated alkane chains has a synergistic effect in improving the binding capacity of the ligands to ERα. The binding modes, SHapley Additive Explanations analysis, and attention map emphasized the importance of π-π stacking and hydrogen bonding interactions with the phenol group, while the fluorinated alkane chain enhanced the interaction with the hydrophobic amino acids of the active pocket. Experimental validation using zebrafish models further confirmed the ERα activity of the representative PFAS molecules. Overall, the current computational workflow is beneficial for the toxicological screening of emerging PFAS and accelerating the development of eco-friendly PFAS molecules, thereby mitigating the environmental and health risks associated with PFAS exposure.

PbImpute: Precise Zero Discrimination and Balanced Imputation in Single-Cell RNA Sequencing Data

Single-cell RNA sequencing (scRNA-seq) has emerged as a transformative technology for elucidating cellular heterogeneity at unprecedented resolution. However, technical limitations such as limited sequencing depth and mRNA capture efficiency often result in zero counts, commonly referred to as "dropout zeros" in scRNA-seq data. These zeros pose significant challenges to downstream analysis, as they can distort the interpretation of cellular transcriptomes. While numerous computational methods have been developed to address this challenge, existing approaches frequently suffer from either insufficient imputation of zeros (under-imputation) or excessive modification of zeros (over-imputation). Here, we propose a precisely balanced imputation (PbImpute) method designed to achieve optimal equilibrium between dropout recovery and biological zero preservation in scRNA-seq data. PbImpute employs a multistage approach: (1) Initial discrimination between technical dropouts and biological zeros through parameter optimization of a new zero-inflated negative binomial (ZINB) distribution model, followed by initial imputation; (2) Application of a uniquely designed static repair algorithm to enhance data fidelity; (3) Secondary dropout identification based on gene expression frequency and partition-specific coefficient of variation; (4) Graph-embedding neural network-based imputation; and (5) Implementation of a uniquely designed dynamic repair mechanism to mitigate over-imputation effects. PbImpute distinguishes itself by uniquely integrating ZINB modeling with static and dynamic repair. This advantageous combined approach achieves a balance between over- and under-imputation, while simultaneously preserving true biological zeros and reducing signal distortion. Comprehensive evaluation using both simulated and real scRNA-seq data sets demonstrated that PbImpute achieves superior performance (F1 Score = 0.88 at 83% dropout rate, ARI = 0.78 on PBMC) in discriminating between technical dropouts and biological zeros compared to state-of-the-art methods. The method significantly improves gene-gene and cell-cell correlation structures, enhances differential expression analysis sensitivity, optimizes clustering resolution and dimensional reduction visualization, and facilitates more accurate trajectory inference. Ablation studies confirmed the essential contribution of both the imputation and repair modules to the method's performance. The code is available at https://github.com/WyBioTeam/PbImpute. By enhancing the accuracy of scRNA-seq data imputation, PbImpute can improve the identification of cell subpopulations and the detection of differentially expressed genes, thereby facilitating more precise analyses of cellular heterogeneity and advancing disease research.

Ventricular Arrhythmia Classification Using Similarity Maps and Hierarchical Multi-Stream Deep Learning

OBJECTIVE

Ventricular arrhythmias are the primary arrhythmias that cause sudden cardiac death. We address the problem of classification between ventricular tachycardia (VT), ventricular fibrillation (VF) and non-ventricular rhythms (NVR).

METHODS

To address the challenging problem of the discrimination between VT and VF, we develop similarity maps - a novel set of features designed to capture regularity within an ECG trace. These similarity maps are combined with features extracted through learnable Parzen band-pass filters and derivative features to discriminate between VT, VF, and NVR. To combine the benefits of these different features, we propose a hierarchical multi-stream ResNet34 architecture.

RESULTS

Our empirical results demonstrate that the similarity maps significantly improve the accuracy of distinguishing between VT and VF. Overall, the proposed approach achieves an average class sensitivity of 89.68%, and individual class sensitivities of 81.46% for VT, 89.29% for VF, and 98.28% for NVR.

CONCLUSION

The proposed method achieves a high accuracy of ventricular arrhythmia detection and classification.

SIGNIFICANCE

Correct detection and classification of ventricular fibrillation and ventricular tachycardia are essential for effective intervention and for the development of new therapies and translational medicine.

The molecular structure of SHISA5 protein and its novel role in primary biliary cholangitis: From single-cell RNA sequencing to biomarkers

The study collected liver tissue samples from PBC patients and healthy controls and performed transcriptomic analysis of the cells in the samples using single-cell RNA sequencing. The expression characteristics of SHISA5 in PBC were revealed by comparing the difference of SHISA5 protein in the two groups of samples. The structure of SHISA5 protein was predicted and its possible biological function was analysed by bioinformatics method. The results showed that the expression of SHISA5 protein in liver tissue of PBC patients was significantly higher than that of healthy controls. Single-cell RNA sequencing data showed that SHISA5 was mainly expressed in hepatocytes and bile duct cells, and its expression level was positively correlated with the disease activity of PBC. Through structural prediction, we found that the SHISA5 protein molecule has a unique transmembrane domain and may be involved in cell signaling and intercellular interactions. Further functional analysis revealed that SHISA5 may participate in the pathological process of PBC by regulating the differentiation and function of bile duct cells.

A novel technology for rapid identification of hemp fibers by terahertz spectroscopy

A novel method for the rapid identification of hemp fibers is proposed in this paper, utilizing terahertz time-domain spectroscopy (THz-TDS) combined with the LargeVis (LV) dimensionality reduction technique. This approach takes advantage of the strengths of THz-TDS while enhancing classification accuracy through LV. To verify the efficacy of this method, terahertz absorption spectral data from three types of hemp fibers were processed. The THz absorption spectra were initially preprocessed using Hanning filtering. Following this, the filtered data underwent dimensionality reduction through three distinct methods: linear Principal Component Analysis (PCA), nonlinear t-Distributed Stochastic Neighbor Embedding (t-SNE), and the LV method. This sequence of steps resulted in a two-dimensional feature data matrix derived from the THz source spectral data. The resultant feature data matrices were then input into both K-Nearest Neighbors (KNN) and Decision Tree (DT) classifiers for analysis. The classification accuracies of six models were evaluated, revealing that the LV-KNN model achieved a 86.67% accuracy rate for the three hemp fiber types. Impressively, the LV-DT model achieved a perfect 100.00% accuracy rate for the same fibers. The LV-DT model, when integrated with THz spectroscopy technology, offers a quick and precise method for identifying various types of hemp fibers. This development introduces an innovative optical measurement scheme for the characterization of textile materials.

NeuTox 2.0: A hybrid deep learning architecture for screening potential neurotoxicity of chemicals based on multimodal feature fusion

Chemically induced neurotoxicity is a critical aspect of chemical safety assessment. Traditional and costly experimental methods call for the development of high-throughput virtual screening. However, the small datasets of neurotoxicity have limited the application of advanced deep learning techniques. The current study developed a hybrid deep learning architecture, NeuTox 2.0, through multimodal feature fusion for enhanced prediction accuracy and generalization ability. We incorporated transfer learning based on self-supervised learning, graph neural networks, and molecular fingerprints/descriptors. Four datasets were used to profile neurotoxicity; these were related to blood-brain barrier permeability, neuronal cytotoxicity, microelectrode array-based neural activity, and mammalian neurotoxicity. Comprehensive performance evaluations demonstrated that NeuTox 2.0 has relatively higher predictive capability across all statistical metrics. Specifically, NeuTox 2.0 exhibits remarkable performance in three of the four datasets. In the BBB dataset, although it does not outperform the PaDEL descriptor model, its performance closely approximates that of the top single-modal model. The ablation experiments indicated that NeuTox 2.0 can learn the deeper structural differences of molecules from various feature extractions and capture complex interactions and mapping relationships between various modalities, thereby improving performance for neurotoxicity prediction. Evaluations of anti-noise ability indicated that NeuTox 2.0 has excellent noise resistance relative to traditional machine learning. We applied the NeuTox 2.0 model to predict the neurotoxicity of 315,790 compounds in the REACH database. The results showed that 701 compounds exhibited potential neurotoxicity in the four neurotoxicity-related predictions. In conclusion, NeuTox 2.0 can be used as an efficient tool for early neurotoxicity screening of environmental chemicals.

Evaluation of Prediction Models for the Capping and Breaking Force of Tablets Using Machine Learning Tools in Wet Granulation Commercial-Scale Pharmaceutical Manufacturing

Background/Objectives: This study aimed to establish a predictive model for critical quality attributes (CQAs) related to tablet integrity, including tablet breaking force (TBF), friability, and capping occurrence, using machine learning-based models and nondestructive experimental data. Methods: The machine learning-based models were trained on data to predict the CQAs of metformin HCl (MF)-containing tablets using a commercial-scale wet granulation process, and five models were each compared for regression and classification. We identified eight input variables associated with the process and material parameters that control the tableting outcome using feature importance analysis. Results: Among the models, the Gaussian Process regression model provided the most successful results, with R2 values of 0.959 and 0.949 for TBF and friability, respectively. Capping occurrence was accurately predicted by all models, with the Boosted Trees model achieving a 97.80% accuracy. Feature importance analysis revealed that the compression force and magnesium stearate fraction were the most influential parameters in CQA prediction and are input variables that could be used in CQA prediction. Conclusions: These findings indicate that TBF, friability, and capping occurrence were successfully modeled using machine learning with a large dataset by constructing regression and classification models. Applying these models before tablet manufacturing can enhance product quality during wet granulation scale-up, particularly by preventing capping during the manufacturing process without damaging the tablets.

Adipose tissue protects against skin photodamage through CD151- and AdipoQ- EVs

To clarify the protective effects of subcutaneous adipose tissue (SAT) against photodamage, we utilized nude mouse skin with or without SAT. Skin and fibroblasts were treated with adipose tissue-derived extracellular vesicles (AT-EVs) or extracellular vesicles derived from adipose-derived stem cells (ADSC-EVs) to demonstrate that SAT protects the overlying skin from photodamage primarily through AT-EVs. Surprisingly, AT-EVs stimulated fibroblast proliferation more rapidly than ADSC-EVs did. The yield of AT-EVs from the same volume of AT was 200 times greater than that of ADSC-EVs. To compare the differences between AT-EVs and ADSC-EVs, we used a proximity barcoding assay (PBA) to analyze the surface proteins on individual particles of these two types of EVs. PBA analysis revealed that AT-EVs contain diverse subpopulations, with 83.42% expressing CD151, compared to only 1.98% of ADSC-EVs. Furthermore, AT-EVs are internalized more rapidly by cells than ADSC-EVs, as our study demonstrated that CD151-positive AT-EVs were endocytosed more quickly than their CD151-negative counterparts. Additionally, adiponectin in AT-EVs activated the AMPK pathway and inhibited the NF-κB pathway, enhancing fibroblast protection against photodamage. The significantly higher yield and faster acquisition of AT-EVs compared to ADSC-EVs underscore their potential for broader applications.

Application of machine learning for mass spectrometry-based multi-omics in thyroid diseases

Thyroid diseases, including functional and neoplastic diseases, bring a huge burden to people's health. Therefore, a timely and accurate diagnosis is necessary. Mass spectrometry (MS) based multi-omics has become an effective strategy to reveal the complex biological mechanisms of thyroid diseases. The exponential growth of biomedical data has promoted the applications of machine learning (ML) techniques to address new challenges in biology and clinical research. In this review, we presented the detailed review of applications of ML for MS-based multi-omics in thyroid disease. It is primarily divided into two sections. In the first section, MS-based multi-omics, primarily proteomics and metabolomics, and their applications in clinical diseases are briefly discussed. In the second section, several commonly used unsupervised learning and supervised algorithms, such as principal component analysis, hierarchical clustering, random forest, and support vector machines are addressed, and the integration of ML techniques with MS-based multi-omics data and its application in thyroid disease diagnosis is explored.

Decoding the mosaic of inflammatory bowel disease: Illuminating insights with single-cell RNA technology

Inflammatory bowel diseases (IBD), comprising ulcerative colitis (UC) and Crohn's disease (CD), are complex chronic inflammatory intestinal conditions with a multifaceted pathology, influenced by immune dysregulation and genetic susceptibility. The challenges in understanding IBD mechanisms and implementing precision medicine include deciphering the contributions of individual immune and non-immune cell populations, pinpointing specific dysregulated genes and pathways, developing predictive models for treatment response, and advancing molecular technologies. Single-cell RNA sequencing (scRNA-seq) has emerged as a powerful tool to address these challenges, offering comprehensive transcriptome profiles of various cell types at the individual cell level in IBD patients, overcoming limitations of bulk RNA sequencing. Additionally, single-cell proteomics analysis, T-cell receptor repertoire analysis, and epigenetic profiling provide a comprehensive view of IBD pathogenesis and personalized therapy. This review summarizes significant advancements in single-cell sequencing technologies for enhancing our understanding of IBD, covering pathogenesis, diagnosis, treatment, and prognosis. Furthermore, we discuss the challenges that persist in the context of IBD research, including the need for longitudinal studies, integration of multiple single-cell and spatial transcriptomics technologies, and the potential of microbial single-cell RNA-seq to shed light on the role of the gut microbiome in IBD.

Molecular designing of potential environmentally friendly PFAS based on deep learning and generative models

Perfluoroalkyl and polyfluoroalkyl substances (PFAS) are widely used across a spectrum of industrial and consumer goods. Nonetheless, their persistent nature and tendency to accumulate in biological systems pose substantial environmental and health threats. Consequently, striking a balance between maximizing product efficiency and minimizing environmental and health risks by tailoring the molecular structure of PFAS has become a pivotal challenge in the fields of environmental chemistry and sustainable development. To address this issue, a computational workflow was proposed for designing an environmentally friendly PFAS by incorporating deep learning (DL) and molecular generative models. The hybrid DL architecture MolHGT+ based on heterogeneous graph neural network with transformer-like attention was applied to predict the surface tension, bioaccumulation, and hepatotoxicity of the molecules. Through virtual screening of the PFAS master database using MolHGT+, the findings indicate that incorporating the siloxane group and betaine fragment can effectively decrease both the bioaccumulation and hepatotoxicity of PFAS while preserving low surface tension. In addition, molecular generative models were employed to create a structurally diverse pool of novel PFASs with the aforementioned hit molecules serving as the initial template structures. Overall, our study presents a promising AI-driven method for advancing the development of environmentally friendly PFAS.

CPARI: a novel approach combining cell partitioning with absolute and relative imputation to address dropout in single-cell RNA-seq data

A key challenge in analyzing single-cell RNA sequencing data is the large number of false zeros, known as "dropout zeros", which are caused by technical limitations such as shallow sequencing depth or inefficient mRNA capture. To address this challenge, we propose a novel imputation model called CPARI, which combines cell partitioning with our designed absolute and relative imputation methods. Initially, CPARI employs a new approach to select highly variable genes and constructs an average consensus matrix using C-mean fuzzy clustering-based blockchain technology to obtain results at different resolutions. Hierarchical clustering is then applied to further refine these blocks, resulting in well-defined cellular partitions. Subsequently, CPARI identifies dropout events and determines the imputation positions of these identified zeros. An autoencoder is trained within each cellular block to learn gene features and reconstruct data. Our uniquely defined absolute imputation technique is first applied to the identified positions, followed by our relative imputation technique to address remaining dropout zeros, ensuring that both global consistency and local variation are maintained. Through comprehensive analyses conducted on simulated and real scRNA-seq datasets, including quantitative assessment, differential expression analysis, cell clustering, cell trajectory inference, robustness evaluation, and large-scale data imputation, CPARI demonstrates superior performance compared to 12 other art-of-state imputation models. Additionally, ablation experiments further confirm the significance and necessity of both the cell partitioning and relative imputation components of CPARI. Notably, CPARI as a new denoising approach could distinguish between real biological zeros and dropout zeros and minimize false positives, and maximize the accuracy of imputation.

EAD: effortless anomalies detection, a deep learning based approach for detecting outliers in English textual data

Anomalies are the existential abnormalities in data, the identification of which is known as anomaly detection. The absence of timely detection of anomalies may affect the key processes of decision-making, fraud detection, and automated classification. Most of the existing models of anomaly detection utilize the traditional way of tokenizing and are computationally costlier, mainly if the outliers are to be extracted from a large script. This research work intends to propose an unsupervised, all-MiniLM-L6-v2-based system for the detection of outliers. The method makes use of centroid embeddings to extract outliers in high-variety, large-volume data. To avoid mistakenly treating novelty as an outlier, the Minimum Covariance Determinant (MCD) based approach is followed to count the novelty of the input script. The proposed method is implemented in a Python project, App. for Anomalies Detection (AAD). The system is evaluated by two non-related datasets-the 20 newsgroups text dataset and the SMS spam collection dataset. The robust accuracy (94%) and F1 score (0.95) revealed that the proposed method could effectively trace anomalies in a comparatively large script. The process is applicable in extracting meanings from textual data, particularly in the domains of human resource management and security.

Combining machine learning, molecular dynamics, and free energy analysis for (5HT)-2A receptor modulator classification

The 5-Hydroxytryptamine (5HT)-2A receptor, a key target in psychoactive drug development, presents significant challenges in the design of selective compounds. Here, we describe the construction, evaluation and validation of two machine learning (ML) models for the classification of bioactivity mechanisms against the (5HT)-2A receptor. Employing neural networks and XGBoost models, we achieved an overall accuracy of around 87 %, which was further enhanced through molecular modelling (MM) (e.g. molecular dynamics simulations) and binding free energy analysis. This ML-MM integration provided insights into the mechanisms of direct modulators and prodrugs. A significant outcome of the current study is the development of a 'binding free energy fingerprint' specific to (5HT)-2A modulators, offering a novel metric for evaluating drug efficacy against this target. Our study demonstrates the prospective of employing a successful workflow combining AI with structural biology, offering a powerful tool for advancing psychoactive drug discovery.

Enhancing Influenza Detection through Integrative Machine Learning and Nasopharyngeal Metabolomic Profiling: A Comprehensive Study

Background/Objectives: Nasal and nasopharyngeal swabs are commonly used for detecting respiratory viruses, including influenza, which significantly alters host cell metabolites. This study aimed to develop a machine learning model to identify biomarkers that differentiate between influenza-positive and -negative cases using clinical metabolomics data. Method: A publicly available dataset of 236 nasopharyngeal samples screened via liquid chromatography-quadrupole time-of-flight (LC/Q-TOF) mass spectrometry was used. Among these, 118 samples tested positive for influenza (40 A H1N1, 39 A H3N2, 39 Influenza B), while 118 were negative controls. A stacking-based model was proposed using the top 20 selected features. Thirteen machine learning models were initially trained, and the top three were combined using predicted probabilities to form a stacking classifier. Results: The ExtraTrees stacking model outperformed other models, achieving 97.08% accuracy. External validation on a prospective cohort of 96 symptomatic individuals (48 positive and 48 negatives for influenza) showed 100% accuracy. SHAP values were used to enhance model explainability. Metabolites such as Pyroglutamic Acid (retention time: 0.81 min, m/z: 84.0447) and its in-source fragment ion (retention time: 0.81 min, m/z: 130.0507) showed minimal impact on influenza-positive cases. On the other hand, metabolites with a retention time of 10.34 min and m/z 106.0865, and a retention time of 8.65 min and m/z 211.1376, demonstrated significant positive contributions. Conclusions: This study highlights the effectiveness of integrating metabolomics data with machine learning for accurate influenza diagnosis. The stacking-based model, combined with SHAP analysis, provided robust performance and insights into key metabolites influencing predictions.

Relevance of the regulation of the brain-placental axis to the nocturnal bottleneck of mammals

INTRODUCTION

Evolutionary theory suggests that the ancestors of all placental animals were nocturnal. Visual perceptive function of mammalian brain has evolved extensively, but nearly 70 % of today's mammals are still nocturnal. While placental influence on brain development is known, if placenta plays a role in the visual perceptive function of mammalian brain remains untested. The present study aims to test this hypothesis.

METHODS

In this study, single-nuclei RNA sequencing was performed to identify genes expressed in the pig placenta and fetal brain, and then compared with the orthologous genes expressed in the placenta and fetal brain cells of mouse. Differential gene expression analysis was performed to identify placental genes regulated differentially between nocturnal and diurnal animals. Phylogenetic modeling was performed to test correlated evolution between placenta type, and the nocturnal or diurnal activity among different mammals.

RESULTS

The results showed that genes differentially regulated in the fetal brain were related to visual perception whereas the placental genes were related to the nocturnal or diurnal activity in placental animals. Phylogenetic modeling of these genes in thirty-four diverse mammalian species showed evidence for evolutionary link between placenta and the nocturnal/diurnal activity in animals.

DISCUSSION

The findings of this study suggest that the placenta plays a role in the evolution of visual perceptive function of brain to shape the nocturnal or diurnal activity of placental animals.

Plasma extracellular vesicles proteomics in meningioma patients

Tumor derived Extracellular vesicles (EVs) in circulating system may contain tumor-specific markers, and EV detection in body fluids could become an important tool for early tumor diagnosis, prognosis assessment. Meningiomas are the most common benign intracranial tumors, few studies have revealed specific protein markers for meningiomas from patients' body fluids. In this study, using proximity labeling technology and non-tumor patient plasma as a control, we detected protein levels of EVs in plasma samples from meningioma patients before and after surgery. Through bioinformatics analysis, we discovered that the levels of EV count and protein count in meningioma patients were significantly higher than those in healthy controls, and were significantly decreased postoperatively. Among EV proteins in meningioma patients, the levels of MUC1, SIGLEC11, E-Cadherin, KIT, and TASCTD2 were found not only significantly elevated than those in healthy controls, but also significantly decreased after tumor resection. Moreover, using publicly available GEO databases, we verified that the mRNA level of MUC1, SIGLEC11, and CDH1 in meningiomas were significantly higher in comparison with normal dura mater tissues. Additionally, by analyzing human meningioma specimens collected in this study, we validated the protein levels of MUC1 and SIGLEC11 were significantly increased in WHO grade 2 meningiomas and were positively correlated with tumor proliferation levels. This study indicates that meningiomas secret EV proteins into circulating system, which may serve as specific markers for diagnosis, malignancy predicting and tumor recurrent assessment.

Intelligent Fault Diagnosis Method for Rotating Machinery Based on Recurrence Binary Plot and DSD-CNN

To tackle the issue of the traditional intelligent diagnostic algorithm's insufficient utilization of correlation characteristics within the time series of fault signals and to meet the challenges of accuracy and computational complexity in rotating machinery fault diagnosis, a novel approach based on a recurrence binary plot (RBP) and a lightweight, deep, separable, dilated convolutional neural network (DSD-CNN) is proposed. Firstly, a recursive encoding method is used to convert the fault vibration signals of rotating machinery into two-dimensional texture images, extracting feature information from the internal structure of the fault signals as the input for the model. Subsequently, leveraging the excellent feature extraction capabilities of a lightweight convolutional neural network embedded with attention modules, the fault diagnosis of rotating machinery is carried out. The experimental results using different datasets demonstrate that the proposed model achieves excellent diagnostic accuracy and computational efficiency. Additionally, compared with other representative fault diagnosis methods, this model shows better anti-noise performance under different noise test data, and it provides a reliable and efficient reference solution for rotating machinery fault-classification tasks.

Treating chronic atrophic gastritis: identifying sub-population based on real-world TCM electronic medical records

Background and Objective

Chronic atrophic gastritis (CAG) is a complex chronic disease caused by multiple factors that frequently occurs disease in the clinic. The worldwide prevalence of CAG is high. Interestingly, clinical CAG patients often present with a variety of symptom phenotypes, which makes it more difficult for clinicians to treat. Therefore, there is an urgent need to improve our understanding of the complexity of the clinical CAG population, obtain more accurate disease subtypes, and explore the relationship between clinical symptoms and medication. Therefore, based on the integrated platform of complex networks and clinical research, we classified the collected patients with CAG according to their different clinical characteristics and conducted correlation analysis on the classification results to identify more accurate disease subtypes to aid in personalized clinical treatment.

Method

Traditional Chinese medicine (TCM) offers an empirical understanding of the clinical subtypes of complicated disorders since TCM therapy is tailored to the patient's symptom profile. We gathered 6,253 TCM clinical electronic medical records (EMRs) from CAG patients and manually annotated, extracted, and preprocessed the data. A shared symptom-patient similarity network (PSN) was created. CAG patient subgroups were established, and their clinical features were determined through enrichment analysis employing community identification methods. Different clinical features of relevant subgroups were correlated based on effectiveness to identify symptom-botanical botanical drugs correspondence. Moreover, network pharmacology was employed to identify possible biological relationships between screened symptoms and medications and to identify various clinical and molecular aspects of the key subtypes using Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway enrichment analysis.

Results

5,132 patients were included in the study: 2,699 males (52.60%) and 2,433 females (47.41%). The population was divided into 176 modules. We selected the first 3 modules (M29, M3, and M0) to illustrate the characteristic phenotypes and genotypes of CAG disease subtypes. The M29 subgroup was characterized by gastric fullness disease and internal syndrome of turbidity and poison. The M3 subgroup was characterized by epigastric pain and disharmony between the liver and stomach. The M0 subgroup was characterized by epigastric pain and dampness-heat syndrome. In symptom analysis, The top symptoms for symptom improvement in all three subgroups were stomach pain, bloating, insomnia, poor appetite, and heartburn. However, the three groups were different. The M29 subgroup was more likely to have stomach distention, anorexia, and palpitations. Citrus medica, Solanum nigrum, Jiangcan, Shan ci mushrooms, and Dillon were the most popular botanical drugs. The M3 subgroup has a higher incidence of yellow urine, a bitter tongue, and stomachaches. Smilax glabra, Cyperus rotundus, Angelica sinensis, Conioselinum anthriscoides, and Paeonia lactiflora were the botanical drugs used. Vomiting, nausea, stomach pain, and appetite loss are common in the M0 subgroup. The primary medications are Scutellaria baicalensis, Smilax glabra, Picrorhiza kurroa, Lilium lancifolium, and Artemisia scoparia. Through GO and KEGG pathway analysis, We found that in the M29 subgroup, Citrus medica, Solanum nigrum, Jiangcan, Shan ci mushrooms, and Dillon may exert their therapeutic effects on the symptoms of gastric distension, anorexia, and palpitations by modulating apoptosis and NF-κB signaling pathways. In the M3 subgroup, Smilax glabra, Cyperus rotundus, Angelica sinensis, Conioselinum anthriscoides, and Paeonia lactiflora may be treated by NF-κB and JAK-STAT signaling pathway for the treatment of stomach pain, bitter mouth, and yellow urine. In the M0 subgroup, Scutellaria baicalensis, Smilax glabra, Picrorhiza kurroa, Lilium lancifolium, and Artemisia scoparia may exert their therapeutic effects on poor appetite, stomach pain, vomiting, and nausea through the PI3K-Akt signaling pathway.

Conclusion

Based on PSN identification and community detection analysis, CAG population division can provide useful recommendations for clinical CAG treatment. This method is useful for CAG illness classification and genotyping investigations and can be used for other complicated chronic diseases.

Automatic Quality Assessment of Pork Belly via Deep Learning and Ultrasound Imaging

Pork belly, prized for its unique flavor and texture, is often overlooked in breeding programs that prioritize lean meat production. The quality of pork belly is determined by the number and distribution of muscle and fat layers. This study aimed to assess the number of pork belly layers using deep learning techniques. Initially, semantic segmentation was considered, but the intersection over union (IoU) scores for the segmented parts were below 70%, which is insufficient for practical application. Consequently, the focus shifted to image classification methods. Based on the number of fat and muscle layers, a dataset was categorized into three groups: three layers (n = 1811), five layers (n = 1294), and seven layers (n = 879). Drawing upon established model architectures, the initial model was refined for the task of learning and predicting layer traits from B-ultrasound images of pork belly. After a thorough evaluation of various performance metrics, the ResNet18 model emerged as the most effective, achieving a remarkable training set accuracy of 99.99% and a validation set accuracy of 96.22%, with corresponding loss values of 0.1478 and 0.1976. The robustness of the model was confirmed through three interpretable analysis methods, including grad-CAM, ensuring its reliability. Furthermore, the model was successfully deployed in a local setting to process B-ultrasound video frames in real time, consistently identifying the pork belly layer count with a confidence level exceeding 70%. By employing a scoring system with 100 points as the threshold, the number of pork belly layers in vivo was categorized into superior and inferior grades. This innovative system offers immediate decision-making support for breeding determinations and presents a highly efficient and precise method for assessment of pork belly layers.

Macrophage-derived CD36 + exosome subpopulations as novel biomarkers of Candida albicans infection

Invasive candidiasis (IC) is a notable healthcare-associated fungal infection, characterized by high morbidity, mortality, and substantial treatment costs. Candida albicans emerges as a principal pathogen in this context. Recent academic advancements have shed light on the critical role of exosomes in key biological processes, such as immune responses and antigen presentation. This burgeoning body of research underscores the potential of exosomes in the realm of medical diagnostics and therapeutics, particularly in relation to fungal infections like IC. The exploration of exosomal functions in the pathophysiology of IC not only enhances our understanding of the disease but also opens new avenues for innovative therapeutic interventions. In this investigation, we focus on exosomes (Exos) secreted by macrophages, both uninfected and those infected with C. albicans. Our objective is to extract and analyze these exosomes, delving into the nuances of their protein compositions and subgroups. To achieve this, we employ an innovative technique known as Proximity Barcoding Assay (PBA). This methodology is pivotal in our quest to identify novel biological targets, which could significantly enhance the diagnostic and therapeutic approaches for C. albicans infection. The comparative analysis of exosomal contents from these two distinct cellular states promises to yield insightful data, potentially leading to breakthroughs in understanding and treating this invasive fungal infection. In our study, we analyzed differentially expressed proteins in exosomes from macrophages and C. albicans -infected macrophages, focusing on proteins such as ACE2, CD36, CAV1, LAMP2, CD27, and MPO. We also examined exosome subpopulations, finding a dominant expression of MPO in the most prevalent subgroup, and a distinct expression of CD36 in cluster14. These findings are crucial for understanding the host response to C. albicans and may inform targeted diagnostic and therapeutic approaches. Our study leads us to infer that MPO and CD36 proteins may play roles in the immune escape mechanisms of C. albicans. Additionally, the CD36 exosome subpopulations, identified through our analysis, could serve as potential biomarkers and therapeutic targets for C. albicans infection. This insight opens new avenues for understanding the infection's pathology and developing targeted treatments.

AI-Enhanced evaluation of YouTube content on post-surgical incontinence following pelvic cancer treatment

Background

Several pelvic area cancers exhibit high incidence rates, and their surgical treatment can result in adverse effects such as urinary and fecal incontinence, significantly impacting patients' quality of life. Post-surgery incontinence is a significant concern, with prevalence rates ranging from 25 to 45% for urinary incontinence and 9-68% for fecal incontinence. Cancer survivors are increasingly turning to YouTube as a platform to connect with others, yet caution is warranted as misinformation is prevalent.

Objective

This study aims to evaluate the information quality in YouTube videos about post-surgical incontinence after pelvic area cancer surgery.

Methods

A YouTube search for "Incontinence after cancer surgery" yielded 108 videos, which were subsequently analyzed. To evaluate these videos, several quality assessment tools were utilized, including DISCERN, GQS, JAMA, PEMAT, and MQ-VET. Statistical analyses, such as descriptive statistics and intercorrelation tests, were employed to assess various video attributes, including characteristics, popularity, educational value, quality, and reliability. Also, artificial intelligence techniques like PCA, t-SNE, and UMAP were used for data analysis. HeatMap and Hierarchical Clustering Dendrogram techniques validated the Machine Learning results.

Results

The quality scales presented a high level of correlation one with each other (p < 0.01) and the Artificial Intelligence-based techniques presented clear clustering representations of the dataset samples, which were reinforced by the Heat Map and Hierarchical Clustering Dendrogram.

Conclusions

YouTube videos on "Incontinence after Cancer Surgery" present a "High" quality across multiple scales. The use of AI tools, like PCA, t-SNE, and UMAP, is highlighted for clustering large health datasets, improving data visualization, pattern recognition, and complex healthcare analysis.

Deep Attention Networks With Multi-Temporal Information Fusion for Sleep Apnea Detection

Sleep Apnea (SA) is a prevalent sleep disorder with multifaceted etiologies that can have severe consequences for patients. Diagnosing SA traditionally relies on the in-laboratory polysomnogram (PSG), which records various human physiological activities overnight. SA diagnosis involves manual scoring by qualified physicians. Traditional machine learning methods for SA detection depend on hand-crafted features, making feature selection pivotal for downstream classification tasks. In recent years, deep learning has gained popularity in SA detection due to its capability for automatic feature extraction and superior classification accuracy. This study introduces a Deep Attention Network with Multi-Temporal Information Fusion (DAN-MTIF) for SA detection using single-lead electrocardiogram (ECG) signals. This framework utilizes three 1D convolutional neural network (CNN) blocks to extract features from R-R intervals and R-peak amplitudes using segments of varying lengths. Recognizing that features derived from different temporal scales vary in their contribution to classification, we integrate a multi-head attention module with a self-attention mechanism to learn the weights for each feature vector. Comprehensive experiments and comparisons between two paradigms of classical machine learning approaches and deep learning approaches are conducted. Our experiment results demonstrate that (1) compared with benchmark methods, the proposed DAN-MTIF exhibits excellent performance with 0.9106 accuracy, 0.9396 precision, 0.8470 sensitivity, 0.9588 specificity, and 0.8909 [Formula: see text] score at per-segment level; (2) DAN-MTIF can effectively extract features with a higher degree of discrimination from ECG segments of multiple timescales than those with a single time scale, ensuring a better SA detection performance; (3) the overall performance of deep learning methods is better than the classical machine learning algorithms, highlighting the superior performance of deep learning approaches for SA detection.

Using machine learning algorithms based on patient admission laboratory parameters to predict adverse outcomes in COVID-19 patients

Amidst the global COVID-19 pandemic, the urgent need for timely and precise patient prognosis assessment underscores the significance of leveraging machine learning techniques. In this study, we present a novel predictive model centered on routine clinical laboratory test data to swiftly forecast patient survival outcomes upon admission. Our model integrates feature selection algorithms and binary classification algorithms, optimizing algorithmic selection through meticulous parameter control. Notably, we developed an algorithm coupling Lasso and SVM methodologies, achieving a remarkable area under the ROC curve of 0.9277 with the use of merely 8 clinical laboratory parameters collected upon admission. Our primary contribution lies in the utilization of straightforward laboratory parameters for prognostication, circumventing data processing intricacies, and furnishing clinicians with an expeditious and precise prognostic assessment tool.

Discovery of Covalent Lead Compounds Targeting 3CL Protease with a Lateral Interactions Spiking Neural Network

Covalent drugs exhibit advantages in that noncovalent drugs cannot match, and covalent docking is an important method for screening covalent lead compounds. However, it is difficult for covalent docking to screen covalent compounds on a large scale because covalent docking requires determination of the covalent reaction type of the compound. Here, we propose to use deep learning of a lateral interactions spiking neural network to construct a covalent lead compound screening model to quickly screen covalent lead compounds. We used the 3CL protease (3CL Pro) of SARS-CoV-2 as the screen target and constructed two classification models based on LISNN to predict the covalent binding and inhibitory activity of compounds. The two classification models were trained on the covalent complex data set targeting cysteine (Cys) and the compound inhibitory activity data set targeting 3CL Pro, respected, with good prediction accuracy (ACC > 0.9). We then screened the screening compound library with 6 covalent binding screening models and 12 inhibitory activity screening models. We tested the inhibitory activity of the 32 compounds, and the best compound inhibited SARS-CoV-2 3CL Pro with an IC50 value of 369.5 nM. Further assay implied that dithiothreitol can affect the inhibitory activity of the compound to 3CL Pro, indicating that the compound may covalently bind 3CL Pro. The selectivity test showed that the compound had good target selectivity to 3CL Pro over cathepsin L. These correlation assays can prove the rationality of the covalent lead compound screening model. Finally, covalent docking was performed to demonstrate the binding conformation of the compound with 3CL Pro. The source code can be obtained from the GitHub repository (https://github.com/guzh970630/Screen_Covalent_Compound_by_LISNN).

NeuTox: A weighted ensemble model for screening potential neuronal cytotoxicity of chemicals based on various types of molecular representations

Chemical-induced neurotoxicity has been widely brought into focus in the risk assessment of chemical safety. However, the traditional in vivo animal models to evaluate neurotoxicity are time-consuming and expensive, which cannot completely represent the pathophysiology of neurotoxicity in humans. Cytotoxicity to human neuroblastoma cell line (SH-SY5Y) is commonly used as an alternative to animal testing for the assessment of neurotoxicity, yet it is still not appropriate for high throughput screening of potential neuronal cytotoxicity of chemicals. In this study, we constructed an ensemble prediction model, termed NeuTox, by combining multiple machine learning algorithms with molecular representations based on the weighted score of Particle Swarm Optimization. For the test set, NeuTox shows excellent performance with an accuracy of 0.9064, which are superior to the top-performing individual models. The subsequent experimental verifications reveal that 5,5'-isopropylidenedi-2-biphenylol and 4,4'-cyclo-hexylidenebisphenol exhibited stronger SH-SY5Y-based cytotoxicity compared to bisphenol A, suggesting that NeuTox has good generalization ability in the first-tier assessment of neuronal cytotoxicity of BPA analogs. For ease of use, NeuTox is presented as an online web server that can be freely accessed via http://www.iehneutox-predictor.cn/NeuToxPredict/Predict.

Development and verification of a novel immunogenic cell death-related signature for predicting the prognosis and immune infiltration in triple-negative breast cancer

BACKGROUND

Insufficient understanding of the pathogenesis and tumor immunology of triple-negative breast cancer (TNBC) has limited the development of immunotherapy. The importance of tumor microenvironment (TME) in immunotyping, prognostic assessment and immunotherapy efficacy of cancer has been emphasized, however, potential immunogenic cell death (ICD) related genes function in TME of TNBC has been rarely investigated.

AIMS

To initially explore the role and related mechanisms of ICD in TNBC, especially the role played in the TME of TNBC, and to identify different relevant subtypes based on ICD, and then develop an ICD-related risk score to predict each TNBC patient TME status, prognosis and immunotherapy response.

METHODS AND RESULTS

In this study, we identified distinct ICD-related modification patterns based on 158 TNBC cases in the TCGA-TNBC cohort. We then investigated the possible correlation between ICD-related modification patterns and TME cell infiltration characteristics in TNBC. By using univariate Cox and least absolute shrinkage and selection operator (LASSO) regression analysis, we created a risk scoring system (ICD score) to quantifiably evaluate the impact of ICD-related modification patterns in individual TNBC patient. Two different ICD-related modification patterns were found with significant differences in immune infiltration. Lower ICD score was correlated with higher immune infiltration, tumor mutational burden and significantly enriched in immune-related pathways, indicating a strong ability to activate immune response, which might account for relatively favorable prognosis of TNBC patients and could serve as a predictor to select suitable candidates for immunotherapy. We used two independent cohorts, GSE58812 cohort and Metabric cohort to validate prognosis and immunohistochemistry for preliminary in vitro validation.

CONCLUSION

This study evidenced that the ICD-related modification patterns might exert pivotal roles in the immune infiltration landscape of TNBC and ICD score might act as potential predictors of prognostic assessment and immunotherapy response. This research provides unique insights for individualize immune treatment strategies and promising immunotherapy candidates screening.

Single-Cell Transcriptional Response of the Placenta to the Ablation of Caveolin-1: Insights into the Adaptive Regulation of Brain-Placental Axis in Mice

Caveolin-1 (Cav1) is a major plasma membrane protein that plays important functions in cellular metabolism, proliferation, and senescence. Mice lacking Cav1 show abnormal gene expression in the fetal brain. Though evidence for placental influence on brain development is emerging, whether the ablation of Cav1 affects the regulation of the brain-placental axis remains unexamined. The current study tests the hypothesis that gene expression changes in specific cells of the placenta and the fetal brain are linked to the deregulation of the brain-placental axis in Cav1-null mice. By performing single-nuclei RNA sequencing (snRNA-seq) analyses, we show that the abundance of the extravillious trophoblast (EVT) and stromal cells, but not the cytotrophoblast (CTB) or syncytiotrophoblast (STB), are significantly impacted due to Cav1 ablation in mice. Interestingly, specific genes related to brain development and neurogenesis were significantly differentially expressed in trophoblast cells due to Cav1 deletion. Comparison of single-cell gene expression between the placenta and the fetal brain further showed that specific genes such as plexin A1 (Plxna1), phosphatase and actin regulator 1 (Phactr1) and amyloid precursor-like protein 2 (Aplp2) were differentially expressed between the EVT and STB cells of the placenta, and also, between the radial glia and ependymal cells of the fetal brain. Bulk RNA-seq analysis of the whole placenta and the fetal brain further identified genes differentially expressed in a similar manner between the placenta and the fetal brain due to the absence of Cav1. The deconvolution of reference cell types from the bulk RNA-seq data further showed that the loss of Cav1 impacted the abundance of EVT cells relative to the stromal cells in the placenta, and that of the glia cells relative to the neuronal cells in the fetal brain. Together, the results of this study suggest that the ablation of Cav1 causes deregulated gene expression in specific cell types of the placenta and the fetal brain in mice.

Role of caveolin-1 in metabolic programming of fetal brain

Mice lacking caveolin-1 (Cav1), a key protein of plasma membrane, exhibit brain aging at an early adult stage. Here, integrative analyses of metabolomics, transcriptomics, epigenetics, and single-cell data were performed to test the hypothesis that metabolic deregulation of fetal brain due to the ablation of Cav1 is linked to brain aging in these mice. The results of this study show that lack of Cav1 caused deregulation in the lipid and amino acid metabolism in the fetal brain, and genes associated with these deregulated metabolites were significantly altered in the brain upon aging. Moreover, ablation of Cav1 deregulated several metabolic genes in specific cell types of the fetal brain and impacted DNA methylation of those genes in coordination with mouse epigenetic clock. The findings of this study suggest that the aging program of brain is confounded by metabolic abnormalities in the fetal stage due to the absence of Cav1.

Accurate noise-robust classification of Bacillus species from MALDI-TOF MS spectra using a denoising autoencoder

Bacillus strains are ubiquitous in the environment and are widely used in the microbiological industry as valuable enzyme sources, as well as in agriculture to stimulate plant growth. The Bacillus genus comprises several closely related groups of species. The rapid classification of these remains challenging using existing methods. Techniques based on MALDI-TOF MS data analysis hold significant promise for fast and precise microbial strains classification at both the genus and species levels. In previous work, we proposed a geometric approach to Bacillus strain classification based on mass spectra analysis via the centroid method (CM). One limitation of such methods is the noise in MS spectra. In this study, we used a denoising autoencoder (DAE) to improve bacteria classification accuracy under noisy MS spectra conditions. We employed a denoising autoencoder approach to convert noisy MS spectra into latent variables representing molecular patterns in the original MS data, and the Random Forest method to classify bacterial strains by latent variables. Comparison of the DAE-RF with the CM method using the artificially noisy test samples showed that DAE-RF offers higher noise robustness. Hence, the DAE-RF method could be utilized for noise-robust, fast, and neat classification of Bacillus species according to MALDI-TOF MS data.

Reimagining Carbon Nanomaterial Analysis: Empowering Transfer Learning and Machine Vision in Scanning Electron Microscopy for High-Fidelity Identification

In this report, we propose a novel technique for identifying and analyzing diverse nanoscale carbon allotropes using scanning electron micrographs. By precisely controlling the quenching rates of undercooled molten carbon through laser irradiation, we achieved the formation of microdiamonds, nanodiamonds, and Q-carbon films. However, standard laser irradiation without proper undercooling control leads to the formation of sparsely located diverse carbon polymorphs, hindering their discovery and classification through manual analyses. To address this challenge, we applied transfer-learning approaches using convolutional neural networks and computer vision techniques to achieve allotrope discovery even with sparse spatial presence. Our method achieved high accuracy rates of 92% for Q-carbon identification and 94% for distinguishing it from nanodiamonds. By leveraging scanning electron micrographs and precise undercooling control, our technique enables the efficient identification and characterization of nanoscale carbon structures. This research significantly contributes to the advancement of the field, providing automated tools for Q-materials and carbon polymorph identification. It opens up new opportunities for the further exploration of these materials in various applications.

Ascites-derived CDCP1+ extracellular vesicles subcluster as a novel biomarker and therapeutic target for ovarian cancer

Introduction

Ovarian cancer (OVCA) is one of the most prevalent malignant tumors of the female reproductive system, and its diagnosis is typically accompanied by the production of ascites. Although liquid biopsy has been widely implemented recently, the diagnosis or prognosis of OVCA based on liquid biopsy remains the primary emphasis.

Methods

In this study, using proximity barcoding assay, a technique for analyzing the surface proteins on single extracellular vesicles (EVs). For validation, serum and ascites samples from patients with epithelial ovarian cancer (EOC) were collected, and their levels of CDCP1 was determined by enzyme-linked immunosorbent assay. Tissue chips were prepared to analyze the relationship between different expression levels of CDCP1 and the prognosis of ovarian cancer patients.

Results

We discovered that the CUB domain-containing protein 1+ (CDCP1+) EVs subcluster was higher in the ascites of OVCA patients compared to benign ascites. At the same time, the level of CDCP1 was considerably elevated in the ascites of OVCA patients. The overall survival and disease-free survival of the group with high CDCP1 expression in EOC were significantly lower than those of the group with low expression. In addition, the receiver operating characteristic curve demonstrates that EVs-derived CDCP1 was a biomarker of early response in OVCA ascites.

Discussion

Our findings identified a CDCP1+ EVs subcluster in the ascites of OVCA patients as a possible biomarker for EOC prevention.

Identifying subgroups of patients with type 2 diabetes based on real-world traditional chinese medicine electronic medical records

Introduction: Type 2 diabetes (T2D) is a multifactorial complex chronic disease with a high prevalence worldwide, and Type 2 diabetes patients with different comorbidities often present multiple phenotypes in the clinic. Thus, there is a pressing need to improve understanding of the complexity of the clinical Type 2 diabetes population to help identify more accurate disease subtypes for personalized treatment. Methods: Here, utilizing the traditional Chinese medicine (TCM) clinical electronic medical records (EMRs) of 2137 Type 2 diabetes inpatients, we followed a heterogeneous medical record network (HEMnet) framework to construct heterogeneous medical record networks by integrating the clinical features from the electronic medical records, molecular interaction networks and domain knowledge. Results: Of the 2137 Type 2 diabetes patients, 1347 were male (63.03%), and 790 were female (36.97%). Using the HEMnet method, we obtained eight non-overlapping patient subgroups. For example, in H3, Poria, Astragali Radix, Glycyrrhizae Radix et Rhizoma, Cinnamomi Ramulus, and Liriopes Radix were identified as significant botanical drugs. Cardiovascular diseases (CVDs) were found to be significant comorbidities. Furthermore, enrichment analysis showed that there were six overlapping pathways and eight overlapping Gene Ontology terms among the herbs, comorbidities, and Type 2 diabetes in H3. Discussion: Our results demonstrate that identification of the Type 2 diabetes subgroup based on the HEMnet method can provide important guidance for the clinical use of herbal prescriptions and that this method can be used for other complex diseases.

Biomarker and genomic analyses reveal molecular signatures of non-cardioembolic ischemic stroke

Acute ischemic stroke (AIS) is a major cause of disability and mortality worldwide. Non-cardioembolic ischemic stroke (NCIS), which constitutes the majority of AIS cases, is highly heterogeneous, thus requiring precision medicine treatments. This study aimed to investigate the molecular mechanisms underlying NCIS heterogeneity. We integrated data from the Third China National Stroke Registry, including clinical phenotypes, biomarkers, and whole-genome sequencing data for 7695 patients with NCIS. We identified 30 molecular clusters based on 63 biomarkers and explored the comprehensive landscape of biological heterogeneity and subpopulations in NCIS. Dimensionality reduction revealed fine-scale subpopulation structures associated with specific biomarkers. The subpopulations with biomarkers for inflammation, abnormal liver and kidney function, homocysteine metabolism, lipid metabolism, and gut microbiota metabolism were associated with a high risk of unfavorable clinical outcomes, including stroke recurrence, disability, and mortality. Several genes encoding potential drug targets were identified as putative causal genes that drive the clusters, such as CDK10, ERCC3, and CHEK2. We comprehensively characterized the genetic architecture of these subpopulations, identified their molecular signatures, and revealed the potential of the polybiomarkers and polygenic prediction for assessing clinical outcomes. Our study demonstrates the power of large-scale molecular biomarkers and genomics to understand the underlying biological mechanisms of and advance precision medicine for NCIS.

Unsupervised machine learning framework for discriminating major variants of concern during COVID-19

Due to the high mutation rate of the virus, the COVID-19 pandemic evolved rapidly. Certain variants of the virus, such as Delta and Omicron emerged with altered viral properties leading to severe transmission and death rates. These variants burdened the medical systems worldwide with a major impact to travel, productivity, and the world economy. Unsupervised machine learning methods have the ability to compress, characterize, and visualize unlabelled data. This paper presents a framework that utilizes unsupervised machine learning methods to discriminate and visualize the associations between major COVID-19 variants based on their genome sequences. These methods comprise a combination of selected dimensionality reduction and clustering techniques. The framework processes the RNA sequences by performing a k-mer analysis on the data and further visualises and compares the results using selected dimensionality reduction methods that include principal component analysis (PCA), t-distributed stochastic neighbour embedding (t-SNE), and uniform manifold approximation projection (UMAP). Our framework also employs agglomerative hierarchical clustering to visualize the mutational differences among major variants of concern and country-wise mutational differences for selected variants (Delta and Omicron) using dendrograms. We also provide country-wise mutational differences for selected variants via dendrograms. We find that the proposed framework can effectively distinguish between the major variants and has the potential to identify emerging variants in the future.

Developing four cuproptosis-related lncRNAs signature to predict prognosis and immune activity in ovarian cancer

BACKGROUND

There has been a recent discovery of a new type of cell death produced by copper-iron ions, called Cuproptosis (copper death). The purpose of this study was to identify LncRNA signatures associated with Cuproptosis in ovarian cancer that could be used as prognostic indicators.

METHODS

RNA sequencing (RNA-seq) profiles with clinicopathological data from TCGA database were used to select prognostic CRLs and then constructed prognostic risk model using multivariate regression analysis and LASSO algorithms. An independent dataset from GEO database was used to validate the prognostic performance. Combined with clinical factors, we further constructed a prognostic nomogram. In addition, tumor immune microenvironment, somatic mutation and drug sensitivity were analyzed using ssGSEA, GSVA, ESTIMATE and CIBERSORT algorithms.

RESULT

A total of 129 CRLs were selected whose expression levels were significantly related to expression levels of 10 cuproptosis-related genes. The univariate Cox regression analysis showed that 12 CRLs were associated with overall survival (OS). Using LASSO algorithms and multivariate regression analysis, we constructed a four-CRLs prognostic signature in the training dataset. Patients in the training dataset could be classified into high- or low-risk subgroups with significantly different OS (log-rank p < 0.001). The prognostic performance was confirmed in TCGA-OC cohort (log-rank p < 0.001) and an independent GEO cohort (log-rank p = 0.023). Multivariate cox regression analysis proved the four-CRLs signature was an independent prognostic factor for OC. Additionally, different risk subtypes showed significantly different levels of immune cells, signal pathways, and drug response.

CONCLUSION

We established a prognostic signature based on cuproptosis-related lncRNAs for OC patients, which will be of great value in predicting the prognosis patients and may provide a new perspective for research and individualized treatment.

Identification of telomere-related genes associated with aging-related molecular clusters and the construction of a diagnostic model in Alzheimer's disease based on a bioinformatic analysis

BACKGROUND

Alzheimer's disease (AD) is a neurodegenerative disease that is strongly associated with aging. Telomeres are DNA sequences that protect chromosomes from damage and shorten with age. Telomere-related genes (TRGs) may play a role in AD's pathogenesis.

OBJECTIVES

To identify TRGs related to aging clusters in AD patients, explore their immunological characteristics, and build a TRG-based prediction model for AD and AD subtypes.

METHODS

We analyzed the gene expression profiles of 97 AD samples from the GSE132903 dataset, using aging-related genes (ARGs) as clustering variables. We also assessed immune-cell infiltration in each cluster. We performed a weighted gene co-expression network analysis to identify cluster-specific differentially expressed TRGs. We compared four machine-learning models (random forest, generalized linear model [GLM], gradient boosting model, and support vector machine) for predicting AD and AD subtypes based on TRGs and validated TRGs by conducting an artificial neural network (ANN) analysis and a nomogram model.

RESULTS

We identified two aging clusters in AD patients with distinct immunological features: Cluster A had higher immune scores than Cluster B. Cluster A and the immune system are intimately associated, and this association could affect immunological function and result in AD via the digestive system. The GLM predicted AD and AD subtypes most accurately and was validated by the ANN analysis and nomogram model.

CONCLUSION

Our analyses revealed novel TRGs associated with aging clusters in AD patients and their immunological characteristics. We also developed a promising prediction model based on TRGs for assessing AD risk.

Intermediate-aided allostery mechanism for α-glucosidase by Xanthene-11v as an inhibitor using residue interaction network analysis

Exploring allosteric inhibition and the discovery of new inhibitor binding sites are important studies in protein regulation mechanisms and drug discovery. Structural and network-based analyses of trajectories resulting from molecular dynamics (MD) simulations have been developed to discover protein dynamics, landscape, functions, and allosteric regions. Here, an experimentally suggested non-competitive inhibitor, xanthene-11v, was considered to explore its allosteric inhibition mechanism in α-glucosidase MAL12. Comparative structural and network analyses were applied to eight 250 ns independent MD simulations, four of which were performed in the free state and four of which were performed in ligand-bound forms. Projected two-dimensional free energy landscapes (FEL) were constructed from the probabilistic distribution of conformations along the first two principal components. The post-simulation analyses of the coordinates, side-chain torsion angles, non-covalent interaction networks, network communities, and their centralities were performed on α-glucosidase conformations and the intermediate sub-states. Important communities of residues have been found that connect the allosteric site to the active site. Some of these residues like Thr307, Arg312, TYR344, ILE345, Phe357, Asp406, Val407, Asp408, and Leu436 are the key messengers in the transition pathway between allosteric and active sites. Evaluating the probability distribution of distances between gate residues including Val407 in one community and Phe158, and Pro65 in another community depicted the closure of this gate due to the inhibitor binding. Six macro states of protein were deduced from the topology of FEL and analysis of conformational preference of free and ligand-bound systems to these macro states shows a combination of lock-and-key, conformational selection, and induced fit mechanisms are effective in ligand binding. All these results reveal structural states, allosteric mechanisms, and key players in the inhibition pathway of α-glucosidase by xanthene-11v.

ETSNet: A deep neural network for EEG-based temporal-spatial pattern recognition in psychiatric disorder and emotional distress classification

The use of EEG for evaluating and diagnosing neurological abnormalities related to psychiatric diseases and identifying human emotions has been improved by deep learning advancements. This research aims to categorize individuals with schizophrenia (SZ), their biological relatives (REL), and healthy controls (HC) using resting EEG brain source signal data defined by regions of interest (ROIs). The proposed solution is a deep neural network for the cortical source signals of the ROIs, incorporating a Squeeze-and-Excitation Block and multiple CNNs designed for eyes-open and closed resting states. The model, called EEG Temporal Spatial Network (ETSNet), has two variants: ETSNet^s and ETSNet^f. Two evaluations were conducted to show the effectiveness of the proposed model. The average accuracy for the classification of SZ, REL, and HC using EEG resting data was 99.57% (ETSNet^f), and the average accuracy for the classification of eyes-open (EO) and eyes-closed (EC) resting states was 93.15% (ETSNet^s). An ablation study was also conducted using two public datasets for intellectual and developmental disorders and emotional states, showing improved classification accuracy compared to advanced EEG classification algorithms when using ETSNet^s.

A Unified Deep Learning Framework for Single-Cell ATAC-Seq Analysis Based on ProdDep Transformer Encoder

Recent advances in single-cell sequencing assays for the transposase-accessibility chromatin (scATAC-seq) technique have provided cell-specific chromatin accessibility landscapes of cis-regulatory elements, providing deeper insights into cellular states and dynamics. However, few research efforts have been dedicated to modeling the relationship between regulatory grammars and single-cell chromatin accessibility and incorporating different analysis scenarios of scATAC-seq data into the general framework. To this end, we propose a unified deep learning framework based on the ProdDep Transformer Encoder, dubbed PROTRAIT, for scATAC-seq data analysis. Specifically motivated by the deep language model, PROTRAIT leverages the ProdDep Transformer Encoder to capture the syntax of transcription factor (TF)-DNA binding motifs from scATAC-seq peaks for predicting single-cell chromatin accessibility and learning single-cell embedding. Based on cell embedding, PROTRAIT annotates cell types using the Louvain algorithm. Furthermore, according to the identified likely noises of raw scATAC-seq data, PROTRAIT denoises these values based on predated chromatin accessibility. In addition, PROTRAIT employs differential accessibility analysis to infer TF activity at single-cell and single-nucleotide resolution. Extensive experiments based on the Buenrostro2018 dataset validate the effeteness of PROTRAIT for chromatin accessibility prediction, cell type annotation, and scATAC-seq data denoising, therein outperforming current approaches in terms of different evaluation metrics. Besides, we confirm the consistency between the inferred TF activity and the literature review. We also demonstrate the scalability of PROTRAIT to analyze datasets containing over one million cells.

Bubble: a fast single-cell RNA-seq imputation using an autoencoder constrained by bulk RNA-seq data

Single-cell RNA-sequencing technology (scRNA-seq) brings research to single-cell resolution. However, a major drawback of scRNA-seq is large sparsity, i.e. expressed genes with no reads due to technical noise or limited sequence depth during the scRNA-seq protocol. This phenomenon is also called 'dropout' events, which likely affect downstream analyses such as differential expression analysis, the clustering and visualization of cell subpopulations, cellular trajectory inference, etc. Therefore, there is a need to develop a method to identify and impute these dropout events. We propose Bubble, which first identifies dropout events from all zeros based on expression rate and coefficient of variation of genes within cell subpopulation, and then leverages an autoencoder constrained by bulk RNA-seq data to only impute those values. Unlike other deep learning-based imputation methods, Bubble fuses the matched bulk RNA-seq data as a constraint to reduce the introduction of false positive signals. Using simulated and several real scRNA-seq datasets, we demonstrate that Bubble enhances the recovery of missing values, gene-to-gene and cell-to-cell correlations, and reduces the introduction of false positive signals. Regarding some crucial downstream analyses of scRNA-seq data, Bubble facilitates the identification of differentially expressed genes, improves the performance of clustering and visualization, and aids the construction of cellular trajectory. More importantly, Bubble provides fast and scalable imputation with minimal memory usage.

Structured deep embedding model to generate composite clinical indices from electronic health records for early detection of pancreatic cancer

The high-dimensionality, complexity, and irregularity of electronic health records (EHR) data create significant challenges for both simplified and comprehensive health assessments, prohibiting an efficient extraction of actionable insights by clinicians. If we can provide human decision-makers with a simplified set of interpretable composite indices (i.e., combining information about groups of related measures into single representative values), it will facilitate effective clinical decision-making. In this study, we built a structured deep embedding model aimed at reducing the dimensionality of the input variables by grouping related measurements as determined by domain experts (e.g., clinicians). Our results suggest that composite indices representing liver function may consistently be the most important factor in the early detection of pancreatic cancer (PC). We propose our model as a basis for leveraging deep learning toward developing composite indices from EHR for predicting health outcomes, including but not limited to various cancers, with clinically meaningful interpretations.

Application of GeneCloudOmics: Transcriptomic Data Analytics for Synthetic Biology

Research in synthetic biology and metabolic engineering require a deep understanding on the function and regulation of complex pathway genes. This can be achieved through gene expression profiling which quantifies the transcriptome-wide expression under any condition, such as a cell development stage, mutant, disease, or treatment with a drug. The expression profiling is usually done using high-throughput techniques such as RNA sequencing (RNA-Seq) or microarray. Although both methods are based on different technical approaches, they provide quantitative measures of the expression levels of thousands of genes. The expression levels of the genes are compared under different conditions to identify the differentially expressed genes (DEGs), the genes with different expression levels under different conditions. DEGs, usually involving thousands in number, are then investigated using bioinformatics and data analytic tools to infer and compare their functional roles between conditions. Dealing with such large datasets, therefore, requires intensive data processing and analyses to ensure its quality and produce results that are statistically sound. Thus, there is a need for deep statistical and bioinformatics knowledge to deal with high-throughput gene expression data. This represents a barrier for wet biologists with limited computational, programming, and data analytic skills that prevent them from getting the full potential of the data. In this chapter, we present a step-by-step protocol to perform transcriptome analysis using GeneCloudOmics, a cloud-based web server that provides an end-to-end platform for high-throughput gene expression analysis.

Identification of a novel reactive oxygen species (ROS)-related genes model combined with RT-qPCR experiments for prognosis and immunotherapy in gastric cancer

Reactive oxygen species play a crucial role in the prognosis and tumor microenvironment (TME) of malignant tumors. An ROS-related signature was constructed in gastric cancer (GC) samples from TCGA database. ROS-related genes were obtained from the Molecular Signatures Database. Consensus clustering was used to establish distinct ROS-related subtypes related to different survival and immune cell infiltration patterns. Sequentially, prognostic genes were identified in the ROS-related subtypes, which were used to identify a stable ROS-related signature that predicted the prognosis of GC. Correlation analysis revealed the significance of immune cell iniltration, immunotherapy, and drug sensitivity in gastric cancers with different risks. The putative molecular mechanisms of the different gastric cancer risks were revealed by functional enrichment analysis. A robust nomogram was established to predict the outcome of each gastric cancer. Finally, we verified the expression of the genes involved in the model using RT-qPCR. In conclusion, the ROS-related signature in this study is a novel and stable biomarker associated with TME and immunotherapy responses.

Does Father Christmas Have a Distinctive Facial Phenotype?

We investigated whether Father Christmas has a distinguishable facial phenotype by performing a cross-sectional cohort study examining the facial feature vectors of all publicly available photographs obtained from a google image search of individuals meeting our eligibility criteria presenting as Father Christmas compared with other adult and elderly bearded men. Facial feature vectors were determined using the open-source OpenFace facial recognition system and assessed by support vector machines (SVM). SVM classifiers were trained to distinguish between the facial feature vectors from our groups. Accuracy, precision, and recall results were calculated and the area under the curve (AUC) of the receiver operating characteristic (ROC) were reported for each classifier. SVM classifiers were able to distinguish the face of Father Christmas from other adult men with a high degree of accuracy and could discriminate Father Christmas from elderly bearded men but with lower accuracy. Father Christmas appears to have a distinct facial phenotype when compared to adult men and elderly bearded men. This will be reassuring to children who may be keen to recognise him but raises some interesting questions about the careful use of two-dimensional facial analysis, particularly when employed to explore the relationships between genotype and facial phenotype in a clinical dysmorphology setting.

Distinct tumor microenvironment landscapes of rectal cancer for prognosis and prediction of immunotherapy response

PURPOSE

Tumor microenvironment (TME) affects the progression of rectal cancer (RC), and the clinical relevance of its immune elements was widely reported. Here we aim to delineate the complete TME landscape, including non-immune features, to improve our understanding of RC heterogeneity and provide a better strategy for precision medicine.

METHODS

Single-cell analysis of GSE161277 using Seurat and Cellcall was performed to identify cell-cell interactions. The ssGSEA was employed to quantify the TME elements in TCGA patients, which were further clustered into subtypes by hclust. WGCNA and LASSO were combined to construct a degenerated signature for prognosis, and its performance was validated in two GEO datasets.

RESULTS

We proposed a subtyping strategy based on the abundance of both immune and non-immune components, which divided all RC patients into 4 subtypes (Immune-, Canonical-, Dormant- and Stem-like). Different subtypes exhibited distinct mutation landscapes, biological features, immune characteristics, immunotherapy responses and prognoses. Next, WGCNA and LASSO regression were combined to construct a 10-gene signature based on differentially expressed genes among different subtypes. Subgroups divided by this signature also exhibited different clinical parameters and responses to immune checkpoint blockades. Diverse machine learning algorithms were applied to achieve higher accuracy for survival prediction and a nomogram was further established in combination with M stage and age to provide an accurate and visual prediction of prognosis.

CONCLUSIONS

We identified four TME-based RC subtypes with distinct biological and clinical features. Based on those subtypes, we also proposed a degenerated 10-gene signature to predict the prognosis and immunotherapy response.

Deep learning-based transcriptome model predicts survival of T-cell acute lymphoblastic leukemia

Identifying subgroups of T-cell acute lymphoblastic leukemia (T-ALL) with poor survival will significantly influence patient treatment options and improve patient survival expectations. Current efforts to predict T-ALL survival expectations in multiple patient cohorts are lacking. A deep learning (DL)-based model was developed to determine the prognostic staging of T-ALL patients. We used transcriptome sequencing data from TARGET to build a DL-based survival model using 265 T-ALL patients. We found that patients could be divided into two subgroups (K0 and K1) with significant difference (P&lt; 0.0001) in survival rate. The more malignant subgroup was significantly associated with some tumor-related signaling pathways, such as PI3K-Akt, cGMP-PKG and TGF-beta signaling pathway. DL-based model showed good performance in a cohort of patients from our clinical center (P = 0.0248). T-ALL patients survival was successfully predicted using a DL-based model, and we hope to apply it to clinical practice in the future.

Chemosensory-Related Genes in Marine Copepods

Living organisms deeply rely on the acquisition of chemical signals in any aspect of their life, from searching for food, mating and defending themselves from stressors. Copepods, the most abundant and ubiquitous metazoans on Earth, possess diversified and highly specified chemoreceptive structures along their body. The detection of chemical stimuli activates specific pathways, although this process has so far been analyzed only on a relatively limited number of species. Here, in silico mining of 18 publicly available transcriptomes is performed to delve into the copepod chemosensory genes, improving current knowledge on the diversity of this multigene family and on possible physiological mechanisms involved in the detection and analysis of chemical cues. Our study identifies the presence of ionotropic receptors, chemosensory proteins and gustatory receptors in copepods belonging to the Calanoida, Cyclopoida and Harpacticoida orders. We also confirm the absence in these copepods of odorant receptors and odorant-binding proteins agreeing with their insect specificity. Copepods have evolved several mechanisms to survive in the harsh marine environment such as producing proteins to respond to external stimulii. Overall, the results of our study open new possibilities for the use of the chemosensory genes as biomarkers in chemical ecology studies on copepods and possibly also in other marine holozooplankters.