Wei Zhang

For neurological disorders, single-nucleus RNA sequencing(snRNA-seq) data from human brain samples have revealed valuable insights about regulatory mechanisms that are associated with disease progression. During data mining of RNA-seq data that are associated with Alzheimer's disease(AD) and dementia, conventional deep learning methods generally focus on changes in gene transcript levels, while ignoring graph features of dementia-specific gene networks to a certain degree. It is noted that graph features underlying transcriptomics data have the potential to enhance model performance by analyzing structural information of AD-specific regulatory networks namely AD-GRN. To sufficiently exploit graph features, spatiotemporal graph learning technique has been employed to recognize meaningful patterns that govern AD progression. Using brain snRNA-seq data, this study has developed an ST-GCN architecture, which has embedded a co-attention network and a nonlinear manifold alignment(NMA) fusion block, to explore abnormal regulatory mechanisms about neurological disorders. The co-attention network aims to obtain compact graph representations by compressing evolving AD-GRNs. The proposed STAD-CoAtt method integrates temporal and graph features, thus constructing joint latent representations of snRNA-seq data. Experiments about two benchmark RNA-seq datasets from ROSMAP and GSE platforms have demonstrated the effectiveness and superiority of the STAD-CoAtt method in assessing neuropathology stages and cognitive dysfunction. By investigating cross-view interactions, the proposed STAD-CoAtt method has obtained superior performance over established SOTA approaches in AD classification tasks.

Decoding DNA methylation sites through nanopore sequencing has emerged as a cutting-edge technology in the field of DNA methylation research, as it enables direct sequencing of native DNA molecules without the need for prior enzymatic or chemical treatments. During nanopore sequencing, methylation modifications on DNA bases cause changes in electrical current intensity. Therefore, constructing deep neural network models to decode the electrical signals of nanopore sequencing has become a crucial step in methylation site identification. In this study, we utilized nanopore sequencing data containing diverse DNA methylation types and motif sequence diversity. We proposed a feature encoding method based on current signal clustering and leveraged the powerful attention mechanism in the Transformer framework to construct the PoreFormer model for identifying DNA methylation sites in nanopore sequencing. The model demonstrated excellent performance under conditions of multi-class methylation and motif sequence diversity, offering new insights into related research fields.

Graph learning models have received increasing attention in the computational analysis of single-cell RNA sequencing (scRNA-seq) data. Compared with conventional deep neural networks, graph neural networks and language models have exhibited superior performance by extracting graph-structured data from raw gene count matrices. Established deep neural network-based clustering approaches generally focus on temporal expression patterns while ignoring inherent interactions at gene-level as well as cell-level, which could be regarded as spatial dynamics in single-cell data. Both gene–gene and cell–cell interactions are able to boost the performance of cell type detection, under the framework of multi-view modeling. In this study, spatiotemporal embedding and cell graphs are extracted to capture spatial dynamics at the molecular level. In order to enhance the accuracy of cell type detection, this study proposes the scHybridBERT architecture to conduct multi-view modeling of scRNA-seq data using extracted spatiotemporal patterns. In this scHybridBERT method, graph learning models are employed to deal with cell graphs and the Performer model employs spatiotemporal embeddings. Experimental outcomes about benchmark scRNA-seq datasets indicate that the proposed scHybridBERT method is able to enhance the accuracy of single-cell clustering tasks by integrating spatiotemporal embeddings and cell graphs.

Human blood expression data has played an important role in the study of molecular regulatory mechanisms of diseases. The identification and automatic annotation of cell subtypes can be used to explore disease subtypes and to model the short-term evolution mechanisms of diseases. Due to the weak signal pattern, automatic annotation based on blood expression data is computationally difficult, and effective automatic annotation algorithms are urgently needed to support it. In this paper, human peripheral blood expression data (PBMC) datasets have been selected as benchmark data to validate the performance of cell type detection approaches. In this study, ReconSAGE is proposed to capture the topological structures between cells. This method can better reveal cell subpopulations and improve the accuracy and robustness of cell clustering.

As a kind of terminal neural signal, electromyography (EMG) generated by articulatory muscles is widely used in silent speech recognition (SSR). The SSR task can be converted to a pattern recognition problem, where silent speech patterns correspond to clusters in the feature space. Conventional inductive conformal predictor (ICP) based solutions in the SSR face limitations in the subsampling. In addition, the data quality assessment is a long-standing challenge on the recognition ability of models, but it is seldom discussed in previous works. We introduce the aggregated conformal prediction (ACP) to replace ICP, which enhances the diversity of subsampling to solve the problem. With the purpose of digging out the high-quality extended dataset in data expansion, this study proposes the Supervised K-means Evaluation (SKE) method. Equipped with SKE method, ACP framework contributes to an efficient and robust SSR solution, and the advantages have been validated on the task for Chinese words. Our results support it is significant to design SKE method and utilize ACP framework. To our knowledge, this study is the first to apply ACP methodology and quantitatively evaluate the data quality in the SSR task.

Many effective methods extract and fuse different protein features to study the relationship between protein sequence, structure, and function, but different methods have preferences in solving the research of protein structure and function, which requires selecting valuable and contributing features to design more effective prediction methods. This work mainly focused on the feature selection methods in the study of protein structure and function, and systematically compared and analyzed the efficiency of different feature selection methods in the prediction of protein structures, protein disorders, protein molecular chaperones, and protein solubility. The results show that the feature selection method based on nonlinear SVM performs best in protein structure prediction, protein solubility prediction, protein molecular chaperone prediction, and protein solubility prediction. After selection, the accuracy of features is improved by 13.16% ~71%, especially the Kmer features and PSSM features of proteins.