Interdisciplinary Sciences: Computational Life Sciences最新文献

Stroke is still the World's second major factor of death, as well as the third major factor of death and disability. Ischemic stroke is a type of stroke, in which early detection and treatment are the keys to preventing ischemic strokes. However, due to the limitation of privacy protection and labeling difficulties, there are only a few studies on the intelligent automatic diagnosis of stroke or ischemic stroke, and the results are unsatisfactory. Therefore, we collect some data and propose a 3D carotid Computed Tomography Angiography (CTA) image segmentation model called CA-UNet for fully automated extraction of carotid arteries. We explore the number of down-sampling times applicable to carotid segmentation and design a multi-scale loss function to resolve the loss of detailed features during the process of down-sampling. Moreover, based on CA-Unet, we propose an ischemic stroke risk prediction model to predict the risk in patients using their 3D CTA images, electronic medical records, and medical history. We have validated the efficacy of our segmentation model and prediction model through comparison tests. Our method can provide reliable diagnoses and results that benefit patients and medical professionals.

Single-cell RNA sequencing technology is one of the most cost-effective ways to uncover transcriptomic heterogeneity. With the rapid rise of this technology, enormous amounts of scRNA-seq data have been produced. Due to the high dimensionality, noise, sparsity and missing features of the available scRNA-seq data, accurately clustering the scRNA-seq data for downstream analysis is a significant challenge. Many computational methods have been designed to address this issue; nevertheless, the efficacy of the available methods is still inadequate. In addition, most similarity-based methods require a number of clusters as input, which is difficult to achieve in real applications. In this study, we developed a novel computational method for clustering scRNA-seq data by considering both global and local information, named GCFG. This method characterizes the global properties of data by applying concept factorization, and the regularized Gaussian graphical model is utilized to evaluate the local embedding relationship of data. To learn the cell-cell similarity matrix, we integrated the two components, and an iterative optimization algorithm was developed. The categorization of single cells is obtained by applying Louvain, a modularity-based community discovery algorithm, to the similarity matrix. The behavior of the GCFG approach is assessed on 14 real scRNA-seq datasets in terms of ACC and ARI, and comparison results with 17 other competitive methods suggest that GCFG is effective and robust.

The precise identification of associations between diseases and drugs is paramount for comprehending the etiology and mechanisms underlying parasitic diseases. Computational approaches are highly effective in discovering and predicting disease-drug associations. However, the majority of these approaches primarily rely on link-based methodologies within distinct biomedical bipartite networks. In this study, we reorganized a fundamental dataset of parasitic disease-drug associations using the latest databases, and proposed a prediction model called PDDGCN, based on a multi-view graph convolutional network. To begin with, we fused similarity networks with binary networks to establish multi-view heterogeneous networks. We utilized neighborhood information aggregation layers to refine node embeddings within each view of the multi-view heterogeneous networks, leveraging inter- and intra-domain message passing to aggregate information from neighboring nodes. Subsequently, we integrated multiple embeddings from each view and fed them into the ultimate discriminator. The experimental results demonstrate that PDDGCN outperforms five state-of-the-art methods and four compared machine learning algorithms. Additionally, case studies have substantiated the effectiveness of PDDGCN in identifying associations between parasitic diseases and drugs. In summary, the PDDGCN model has the potential to facilitate the discovery of potential treatments for parasitic diseases and advance our comprehension of the etiology in this field. The source code is available at https://github.com/AhauBioinformatics/PDDGCN .

Circular RNA is a single-stranded RNA with a closed-loop structure. In recent years, academic research has revealed that circular RNAs play critical roles in biological processes and are related to human diseases. The discovery of potential circRNAs as disease biomarkers and drug targets is crucial since it can help diagnose diseases in the early stages and be used to treat people. However, in conventional experimental methods, conducting experiments to detect associations between circular RNAs and diseases is time-consuming and costly. To overcome this problem, various computational methodologies are proposed to extract essential features for both circular RNAs and diseases and predict the associations. Studies showed that computational methods successfully predicted performance and made it possible to detect possible highly related circular RNAs for diseases. This study proposes a deep learning-based circRNA-disease association predictor methodology called DCDA, which uses multiple data sources to create circRNA and disease features and reveal hidden feature codings of a circular RNA-disease pair with a deep autoencoder, then predict the relation score of the pair by a deep neural network. Fivefold cross-validation results on the benchmark dataset showed that our model outperforms state-of-the-art prediction methods in the literature with the AUC score of 0.9794.

MicroRNA (miRNA) serves as a pivotal regulator of numerous cellular processes, and the identification of miRNA-disease associations (MDAs) is crucial for comprehending complex diseases. Recently, graph neural networks (GNN) have made significant advancements in MDA prediction. However, these methods tend to learn one type of node representation from a single heterogeneous network, ignoring the importance of multiple network topologies and node attributes. Here, we propose SMDAP (Sequence hierarchical modeling-based Mirna-Disease Association Prediction framework), a novel GNN-based framework that incorporates multiple network topologies and various node attributes including miRNA seed and full-length sequences to predict potential MDAs. Specifically, SMDAP consists of two types of MDA representation: following a heterogeneous pattern, we construct a transfer learning-like synchronous mutual learning network to learn the first MDA representation in conjunction with the miRNA seed sequence. Meanwhile, following a homogeneous pattern, we design a subgraph-inspired asynchronous multi-scale embedding network to obtain the second MDA representation based on the miRNA full-length sequence. Subsequently, an adaptive fusion approach is designed to combine the two branches such that we can score the MDAs by the downstream classifier and infer novel MDAs. Comprehensive experiments demonstrate that SMDAP integrates the advantages of multiple network topologies and node attributes into two branch representations. Moreover, the area under the receiver operating characteristic curve is 0.9622 on DB1, which is a 5.06% increase from the baselines. The area under the precision–recall curve is 0.9777, which is a 7.33% increase from the baselines. In addition, case studies on three human cancers validated the predictive performance of SMDAP. Overall, SMDAP represents a powerful tool for MDA prediction.

Graphical Abstract

Abstract

Drug repositioning is critical to drug development. Previous drug repositioning methods mainly constructed drug–disease heterogeneous networks to extract drug–disease features. However, these methods faced difficulty when we are using structurally simple models to deal with complex heterogeneous networks. Therefore, in this study, the researchers introduced a drug repositioning method named DRDSA. The method utilizes a deep sparse autoencoder and integrates drug–disease similarities. First, the researchers constructed a drug–disease feature network by incorporating information from drug chemical structure, disease semantic data, and existing known drug–disease associations. Then, we learned the low-dimensional representation of the feature network using a deep sparse autoencoder. Finally, we utilized a deep neural network to make predictions on new drug–disease associations based on the feature representation. The experimental results show that our proposed method has achieved optimal results on all four benchmark datasets, especially on the CTD dataset where AUC and AUPR reached 0.9619 and 0.9676, respectively, outperforming other baseline methods. In the case study, the researchers predicted the top ten antiviral drugs for COVID-19. Remarkably, six out of these predictions were subsequently validated by other literature sources.

Abstract

Since the identification of microRNAs (miRNAs), empirical research has demonstrated their crucial involvement in the functioning of organisms. Investigating miRNAs significantly bolsters efforts related to averting, diagnosing, and treating intricate human maladies. Yet, exploring every conceivable miRNA–disease association consumes significant resources and time within conventional wet experiments. On the computational front, forecasting potential miRNA–disease connections serves as a valuable source of preliminary insights for medical investigators. As a result, we have developed a novel matrix factorization model known as Hessian-regularized (L_{2,1}) nonnegative matrix factorization in combination with deep learning for predicting associations between miRNAs and diseases, denoted as (HRL_{2,1})-NMF-DF. In particular, we introduce a novel iterative fusion approach to integrate all similarities. This method effectively diminishes the sparsity of the initial miRNA–disease associations matrix. Additionally, we devise a mixed model framework that utilizes deep learning, matrix decomposition, and singular value decomposition to capture and depict the intricate nonlinear features of miRNA and disease. The prediction performance of the six matrix factorization methods is improved by comparison and analysis, similarity matrix fusion, data preprocessing, and parameter adjustment. The AUC and AUPR obtained by the new matrix factorization model under fivefold cross validation are comparative or better with other matrix factorization models. Finally, we select three diseases including lung tumor, bladder tumor and breast tumor for case analysis, and further extend the matrix factorization model based on deep learning. The results show that the hybrid algorithm combining matrix factorization with deep learning proposed in this paper can predict miRNAs related to different diseases with high accuracy.

Graphical abstract

Gathering information from multi-perspective graphs is an essential issue for many applications especially for protein-ligand-binding affinity prediction. Most of traditional approaches obtained such information individually with low interpretability. In this paper, we harness the rich information from multi-perspective graphs with a general model, which abstractly represents protein-ligand complexes with better interpretability while achieving excellent predictive performance. In addition, we specially analyze the protein-ligand-binding affinity problem, taking into account the heterogeneity of proteins and ligands. Experimental evaluations demonstrate the effectiveness of our data representation strategy on public datasets by fusing information from different perspectives. All codes are available in the https://github.com/Jthy-af/HaPPy .

Classification of glomerular pathology based on histology sections is the key to diagnose the type and degree of kidney diseases. To address problems in the classification of glomerular lesions in children, a deep learning-based complete glomerular classification framework was designed to detect and classify glomerular pathology. A neural network integrating Resnet and Senet (RS-INet) was proposed and a glomerular classification algorithm implemented to achieve high-precision classification of glomerular pathology. SE-Resnet was applied with improvement by transforming the convolutional layer of the original Resnet residual block into a convolutional block with smaller parameters as well as reduced network parameters on the premise of ensuring network performance. Experimental results showed that our algorithm had the best performance in differentiating mesangial proliferative glomerulonephritis (MsPGN), crescent glomerulonephritis (CGN), and glomerulosclerosis (GS) from normal glomerulus (Normal) compared with other classification algorithms. The accuracy rates were 0.960, 0.940, 0.937, and 0.968, respectively. This suggests that the classification algorithm proposed in the present study is able to identify glomerular lesions with a higher precision, and distinguish similar glomerular pathologies from each other.

In view of the major depressive disorder characteristics such as high mortality as well as high recurrence, it is important to explore an objective and effective detection method for major depressive disorder. Considering the advantages complementary of different machine learning algorithms in information mining process, as well as the fusion complementary of different information, in this study, the spatial-temporal electroencephalography fusion framework using neural network is proposed for major depressive disorder detection. Since electroencephalography is a typical time series signal, we introduce recurrent neural network embedded in long short-term memory unit for extract temporal domain features to solve the problem of long-distance information dependence. To reduce the volume conductor effect, the temporal electroencephalography data are mapping into a spatial brain functional network using phase lag index, then the spatial domain features were extracted from brain functional network using 2D convolutional neural networks. Considering the complementarity between different types of features, the spatial-temporal electroencephalography features are fused to achieve data diversity. The experimental results show that spatial-temporal features fusion can improve the detection accuracy of major depressive disorder with a highest of 96.33%. In addition, our research also found that theta, alpha, and full frequency band in brain regions of left frontal, left central, right temporal are closely related to MDD detection, especially theta frequency band in left frontal region. Only using single-dimension EEG data as decision basis, it is difficult to fully explore the valuable information hidden in the data, which affects the overall detection performance of MDD. Meanwhile, different algorithms have their own advantages for different application scenarios. Ideally, different algorithms should use their respective advantages to jointly address complex problems in engineering fields. To this end, we propose a computer-aided MDD detection framework based on spatial-temporal EEG fusion using neural network, as shown in Fig. 1. The simplified process is as follows: (1) Raw EEG data acquisition and preprocessing. (2) The time series EEG data of each channel are input as recurrent neural network (RNN), and RNN is used to process and extract temporal domain (TD) features. (3) The BFN among different EEG channels is constructed, and CNN is used to process and extract the spatial domain (SD) features of the BFN. (4) Based on the theory of information complementarity, the spatial-temporal information is fused to realize efficient MDD detection. Fig. 1 MDD detection framework based on spatial-temporal EEG fusion.