Search

Abstract Background Accurate and robust drug response prediction is of utmost importance in precision medicine. Although many models have been developed to utilize the representations of drugs and cancer cell lines for predicting cancer drug responses (CDR), their performances can be improved by addressing issues such as insufficient data modality, suboptimal fusion algorithms, and poor generalizability for novel drugs or cell lines. Results We introduce TransCDR, which uses transfer learning to learn drug representations and fuses multi-modality features of drugs and cell lines by a self-attention mechanism, to predict the IC 50 values or sensitive states of drugs on cell lines. We are the first to systematically evaluate the generalization of the CDR prediction model to novel (i.e., never-before-seen) compound scaffolds and cell line clusters. TransCDR shows better generalizability than 8 state-of-the-art models. TransCDR outperforms its 5 variants that train drug encoders (i.e., RNN and AttentiveFP) from scratch under various scenarios. The most critical contributors among multiple drug notations and omics profiles are Extended Connectivity Fingerprint and genetic mutation. Additionally, the attention-based fusion module further enhances the predictive performance of TransCDR. TransCDR, trained on the GDSC dataset, demonstrates strong predictive performance on the external testing set CCLE. It is also utilized to predict missing CDRs on GDSC. Moreover, we investigate the biological mechanisms underlying drug response by classifying 7675 patients from TCGA into drug-sensitive or drug-resistant groups, followed by a Gene Set Enrichment Analysis. Conclusions TransCDR emerges as a potent tool with significant potential in drug response prediction.

Abstract This study explores the prediction of mechanical characteristics of linear polyethylene based on oven residence time, employing various regression models and hyper-parameter tuning through the Whale Optimization Algorithm. The dataset comprises one input variable (oven residence time) and three output parameters (Tensile Strength, Impact Strength, and Flexure Strength). The models investigated include Multilayer Perceptron, K-Nearest Neighbors, Support Vector Regression, Polynomial Regression, and Theil–Sen Regression. The results showcased distinct performances across the models for each output parameter. The Polynomial Regression (WOA-PR) method has been identified as the most suitable option for predicting Tensile Strength due to its ability to achieve the lowest errors in terms of Mean Absolute Error, Root Mean Square Error, and Average Absolute Relative Deviation. K-Nearest Neighbors (WOA-KNN) outperforms other models in predicting Impact Strength due to its superior accuracy and reliability. Additionally, Support Vector Regression (WOA-SVR) emerges as the best model for predicting Flexure Strength, showcasing notable performance in minimizing prediction errors. These findings underscore the significance of model selection and optimization techniques in accurately predicting the mechanical properties of polymers, paving the way for enhanced manufacturing processes and material design.

Abstract Light encodes multidimensional information, such as intensity, polarization, and spectrum. Traditional extraction of this light information requires discrete optical components by subdividing the detection area into many “one-to-one” functional pixels. The broadband photodetection of high-dimensional optical information with a single integrated on-chip detector is highly sought after, yet it poses significant challenges. In this study, we employ a metasurface-assisted graphene photodetector, enabling to simultaneously detect and differentiate various polarization states and wavelengths of broadband light (1-8 μm) at the wavelength prediction accuracy of 0.5 μm. The bipolar polarizability empowered by this design allows to decouple multidimensional information (encompassing polarization and wavelength), which can be achieved by encoding vectorial photocurrents with varying polarities and amplitudes. Furthermore, cooperative multiport metasurfaces are adopted and boosted by machine learning techniques. It enables precise spin-wavelength differentiation over an extremely broad wavelength range (1-8 μm). Our innovation offers a recipe for highly compact and high-dimensional spectral-polarization co-detection.

Abstract Background The internal heterogeneity of breast cancer, notably the tumor microenvironment (TME) consisting of malignant and non-malignant cells, has been extensively explored in recent years. The cells in this complex cellular ecosystem activate or suppress tumor immunity through phenotypic changes, secretion of metabolites and cell-cell communication networks. Macrophages, as the most abundant immune cells within the TME, are recruited by malignant cells and undergo phenotypic remodeling. Tumor-associated macrophages (TAMs) exhibit a variety of subtypes and functions, playing significant roles in impacting tumor immunity. However, their precise subtype delineation and specific function remain inadequately defined. Methods The publicly available single-cell transcriptomes of 49,141 cells from eight breast cancer patients with different molecular subtypes and stages were incorporated into our study. Unsupervised clustering and manual cell annotation were employed to accurately classify TAM subtypes. We then conducted functional analysis and constructed a developmental trajectory for TAM subtypes. Subsequently, the roles of TAM subtypes in cell-cell communication networks within the TME were explored using endothelial cells (ECs) and T cells as key nodes. Finally, analyses were repeated in another independent publish scRNA datasets to validate our findings for TAM characterization. Results TAMs are accurately classified into 7 subtypes, displaying anti-tumor or pro-tumor roles. For the first time, we identified a new TAM subtype capable of proliferation and expansion in breast cancer-TUBA1B+ TAMs playing a crucial role in TAMs diversity and tumor progression. The developmental trajectory illustrates how TAMs are remodeled within the TME and undergo phenotypic and functional changes, with TUBA1B+ TAMs at the initial point. Notably, the predominant TAM subtypes varied across different molecular subtypes and stages of breast cancer. Additionally, our research on cell-cell communication networks shows that TAMs exert effects by directly modulating intrinsic immunity, indirectly regulating adaptive immunity through T cells, as well as influencing tumor angiogenesis and lymphangiogenesis through ECs. Conclusions Our study establishes a precise single-cell atlas of breast cancer TAMs, shedding light on their multifaceted roles in tumor biology and providing resources for targeting TAMs in breast cancer immunotherapy.

ObjectivesThis paper investigates the impact of shared lines with different materials on the dynamic response of floating wind farms in order to design a rational shared mooring system. MethodsFirst, shared mooring and traditional (without shared cables) mooring systems are designed for floating wind farms. The dynamic response and economic cost of the two mooring systems are then compared to validate the feasibility of the shared mooring system design. Second, the influence of the two types of shared mooring systems on the mooring response of floating wind farms is studied. Finally, an analysis is made of the influence of load direction and line material on the tension of the shared lines.ResultsThe results indicate that floating wind farms with shared mooring systems using different materials have nearly identical blade tension. However, there are differences in the impact of the two types of shared mooring systems on wind turbine displacement and the mean tension of mooring lines. In addition, the environmental load direction and line material affect wind turbine displacement and shared line tension.ConclusionsThis study provides useful references for selecting material for shared lines in the mooring systems of floating wind farms.

Abstract Background Decoding human genomic sequences requires comprehensive analysis of DNA sequence functionality. Through computational and experimental approaches, researchers have studied the genotype-phenotype relationship and generate important datasets that help unravel complicated genetic blueprints. Thus, the recently developed artificial intelligence methods can be used to interpret the functions of those DNA sequences. Methods This study explores the use of deep learning, particularly pre-trained genomic models like DNA_bert_6 and human_gpt2-v1, in interpreting and representing human genome sequences. Initially, we meticulously constructed multiple datasets linking genotypes and phenotypes to fine-tune those models for precise DNA sequence classification. Additionally, we evaluate the influence of sequence length on classification results and analyze the impact of feature extraction in the hidden layers of our model using the HERV dataset. To enhance our understanding of phenotype-specific patterns recognized by the model, we perform enrichment, pathogenicity and conservation analyzes of specific motifs in the human endogenous retrovirus (HERV) sequence with high average local representation weight (ALRW) scores. Results We have constructed multiple genotype-phenotype datasets displaying commendable classification performance in comparison with random genomic sequences, particularly in the HERV dataset, which achieved binary and multi-classification accuracies and F1 values exceeding 0.935 and 0.888, respectively. Notably, the fine-tuning of the HERV dataset not only improved our ability to identify and distinguish diverse information types within DNA sequences but also successfully identified specific motifs associated with neurological disorders and cancers in regions with high ALRW scores. Subsequent analysis of these motifs shed light on the adaptive responses of species to environmental pressures and their co-evolution with pathogens. Conclusions These findings highlight the potential of pre-trained genomic models in learning DNA sequence representations, particularly when utilizing the HERV dataset, and provide valuable insights for future research endeavors. This study represents an innovative strategy that combines pre-trained genomic model representations with classical methods for analyzing the functionality of genome sequences, thereby promoting cross-fertilization between genomics and artificial intelligence.

Abstract Background There is little evidence to comprehensively summarize the adverse events (AEs) profile of intermittent fasting (IF) despite its widespread use in patients with overweight or obesity. Methods We searched the main electronic databases and registry websites to identify eligible randomized controlled trials (RCTs) comparing IF versus control groups. A direct meta-analysis using a fixed-effect model was conducted to pool the risk differences regarding common AEs and dropouts. Study quality was assessed by using the Jadad scale. Pre-specified subgroup and sensitivity analyses were conducted to explore potential heterogeneity. Results A total of 15 RCTs involving 1,365 adult individuals were included. Findings did not show a significant difference between IF and Control in risk rate of fatigue [0%, 95% confidence interval (CI), -1% to 2%; P = 0.61], headache [0%, 95%CI: -1% to 2%; P = 0.86] and dropout [1%, 95%CI: -2% to 4%; P = 0.51]. However, a numerically higher risk of dizziness was noted among the IF alone subgroup with non-early time restricted eating [3%, 95%CI: -0% to 6%; P = 0.08]. Conclusions This meta-analysis suggested that IF was not associated with a greater risk of AEs in adult patients affected by overweight or obesity. Additional large-scale RCTs stratified by key confounders and designed to evaluate the long-term effects of various IF regimens are needed to ascertain these AEs profile.

Background Subclinical insomnia (sINSO) represents an early stage of insomnia but lacks effective biomarkers for its recognition. The electroencephalogram(EEG) microstates, reflecting brain network dynamics, may provide potential biomarkers by comparing resting-state EEG parameters between sINSO patients and healthy controls.Methods Resting-state EEG data from 20 sINSO subjects and 20 healthy controls, under both open and closed eye conditions, were analyzed using microstate clustering (labeled A, B, C, and D) and machine learning to evaluate their discriminative power.Results The microstate global explained variance of the eyes-closed data was better than that of the eyes-open data. In the sINSO group under closed-eye conditions, the tendencies and transition probabilities for microstate changes are as follows: A to D at 7.7%, B to D at 10.7%, C to A at 7.3%, and D to B at 10.8%. Under open-eye conditions, they are: A to C at 9.1%, B to C at 8.4%, C to D at 9.4%, and D to C at 8.9%. Machine learning classification showed higher accuracy in closed-eye conditions, reaching 77.6%.Conclusion Resting-state EEG microstates exhibit significant differences between sINSO and healthy individuals. These microstates are promising biomarkers for distinguishing sINSO, with closed-eye data providing the most reliable discrimination.

Explaining biodiversity is a fundamental issue in ecology. A long-standing puzzle lies in the paradox of the plankton: many species of plankton feeding on a limited variety of resources coexist, apparently flouting the competitive exclusion principle (CEP), which holds that the number of predator (consumer) species cannot exceed that of the resources at a steady state. Here, we present a mechanistic model and demonstrate that intraspecific interference among the consumers enables a plethora of consumer species to coexist at constant population densities with only one or a handful of resource species. This facilitated biodiversity is resistant to stochasticity, either with the stochastic simulation algorithm or individual-based modeling. Our model naturally explains the classical experiments that invalidate the CEP, quantitatively illustrates the universal S-shaped pattern of the rank-abundance curves across a wide range of ecological communities, and can be broadly used to resolve the mystery of biodiversity in many natural ecosystems.

The coupling of railway power conditioner (RPC), network, and train will change the impedance characteristics of the existing traction power supply systems, which will lead to the unclear law of high-frequency harmonic resonance. This paper aims to study the resonance mechanism of the RPC-network train coupling system, and then solve the resonance problem of the existing traction power supply system from the perspective of reshaping the high-frequency impedance of the system. Firstly, the RPC-network-train coupling system’s dynamic node admittance matrix is established. After that, a comprehensive resonance evaluation method is proposed to obtain the resonance impact factors (RIFs) of multiple virtual impedances in the control strategy and the filtering parameters, which can realize the comparison and selection of five existing impedance reshaping strategies. To ensure the safe and stable operation of the RPC-network-train coupling system, a multi-objective sorting optimization strategy (MOSOS) considering resonance suppression, harmonics filtering, filter cost, and control system stability is proposed for the RPC based on the RIFs. Finally, the effectiveness of the proposed method is verified by the simulation result in MATLAB/Simulink and the experiment result in the StarSim hardware-in-the-loop (HIL) platform respectively.

Abstract Vitiligo is a hypopigmented skin disease characterized by the loss of melanin. The progressive nature and widespread incidence of vitiligo necessitate timely and accurate detection. Usually, a single diagnostic test often falls short of providing definitive confirmation of the condition, necessitating the assessment by dermatologists who specialize in vitiligo. However, the current scarcity of such specialized medical professionals presents a significant challenge. To mitigate this issue and enhance diagnostic accuracy, it is essential to build deep learning models that can support and expedite the detection process. This study endeavors to establish a deep learning framework to enhance the diagnostic accuracy of vitiligo. To this end, a comparative analysis of five models including ResNet (ResNet34, ResNet50, and ResNet101 models) and Swin Transformer series (Swin Transformer Base, and Swin Transformer Large models), were conducted under the uniform condition to identify the model with superior classification capabilities. Moreover, the study sought to augment the interpretability of these models by selecting one that not only provides accurate diagnostic outcomes but also offers visual cues highlighting the regions pertinent to vitiligo. The empirical findings reveal that the Swin Transformer Large model achieved the best performance in classification, whose AUC, accuracy, sensitivity, and specificity are 0.94, 93.82%, 94.02%, and 93.5%, respectively. In terms of interpretability, the highlighted regions in the class activation map correspond to the lesion regions of the vitiligo images, which shows that it effectively indicates the specific category regions associated with the decision-making of dermatological diagnosis. Additionally, the visualization of feature maps generated in the middle layer of the deep learning model provides insights into the internal mechanisms of the model, which is valuable for improving the interpretability of the model, tuning performance, and enhancing clinical applicability. The outcomes of this study underscore the significant potential of deep learning models to revolutionize medical diagnosis by improving diagnostic accuracy and operational efficiency. The research highlights the necessity for ongoing exploration in this domain to fully leverage the capabilities of deep learning technologies in medical diagnostics.

Abstract Over the years, microbiome research has achieved tremendous advancements driven by culture‐independent meta‐omics approaches. Despite extensive research, our understanding of the functional roles and causal effects of the microbiome on phenotypes remains limited. In this study, we focused on the rumen metaproteome, combining it with metatranscriptome and metabolome data to accurately identify the active functional distributions of rumen microorganisms and specific functional groups that influence feed efficiency. By integrating host genetics data, we established the potentially causal relationships between microbes‐proteins/metabolites‐phenotype, and identified specific patterns in which functional groups of rumen microorganisms influence host feed efficiency. We found a causal link between Selenomonas bovis and rumen carbohydrate metabolism, potentially mediated by bacterial chemotaxis and a two‐component regulatory system, impacting feed utilization efficiency of dairy cows. Our study on the nutrient utilization functional groups in the rumen of high‐feed‐efficiency dairy cows, along with the identification of key microbiota functional proteins and their potentially causal relationships, will help move from correlation to causation in rumen microbiome research. This will ultimately enable precise regulation of the rumen microbiota for optimized ruminant production.

Introduction: Magnetic Resonance Imaging (MRI) is essential in diagnosing cervical spondylosis, providing detailed visualization of osseous and soft tissue structures in the cervical spine. However, manual measurements hinder the assessment of cervical spine sagittal balance, leading to time-consuming and error-prone processes. This study presents the Pyramid DBSCAN Simple Linear Iterative Cluster (PDB-SLIC), an automated segmentation algorithm for vertebral bodies in T2-weighted MR images, aiming to streamline sagittal balance assessment for spinal surgeons.Method: PDB-SLIC combines the SLIC superpixel segmentation algorithm with DBSCAN clustering and underwent rigorous testing using an extensive dataset of T2-weighted mid-sagittal MR images from 4,258 patients across ten hospitals in China. The efficacy of PDB-SLIC was compared against other algorithms and networks in terms of superpixel segmentation quality and vertebral body segmentation accuracy. Validation included a comparative analysis of manual and automated measurements of cervical sagittal parameters and scrutiny of PDB-SLIC’s measurement stability across diverse hospital settings and MR scanning machines.Result: PDB-SLIC outperforms other algorithms in vertebral body segmentation quality, with high accuracy, recall, and Jaccard index. Minimal error deviation was observed compared to manual measurements, with correlation coefficients exceeding 95%. PDB-SLIC demonstrated commendable performance in processing cervical spine T2-weighted MR images from various hospital settings, MRI machines, and patient demographics.Discussion: The PDB-SLIC algorithm emerges as an accurate, objective, and efficient tool for evaluating cervical spine sagittal balance, providing valuable assistance to spinal surgeons in preoperative assessment, surgical strategy formulation, and prognostic inference. Additionally, it facilitates comprehensive measurement of sagittal balance parameters across diverse patient cohorts, contributing to the establishment of normative standards for cervical spine MR imaging.

Objectives: Clinical molecular genetic testing and molecular imaging dramatically increase the quantity of clinical data. Combined with the extensive application of electronic health records, a medical data ecosystem is forming, which calls for big-data-based medicine models. We tried to use big data analytics to search for similar patients in a cancer cohort, showing how to apply artificial intelligence (AI) algorithms to clinical data processing to obtain clinically significant results, with the ultimate goal of improving healthcare management. Methods: In order to overcome the weaknesses of most data processing algorithms that rely on expert labeling and annotation, we uniformly adopted one-hot encoding for all types of clinical data, calculating the Euclidean distance to measure patient similarity and subgrouping via an unsupervised learning model. Overall survival (OS) was investigated to assess the clinical validity and clinical relevance of the model. Results: We took gastric cancers (GCs) as an example to build a high-dimensional clinical patient similarity network (cPSN). When performing the survival analysis, we found that Cluster_2 had the longest survival rates, while Cluster_5 had the worst prognosis among all the subgroups. As patients in the same subgroup share some clinical characteristics, the clinical feature analysis found that Cluster_2 harbored more lower distal GCs than upper proximal GCs, shedding light on the debates. Conclusion: Overall, we constructed a cancer-specific cPSN with excellent interpretability and clinical significance, which would recapitulate patient similarity in the real-world. The constructed cPSN model is scalable, generalizable, and performs well for various data types.