The protein inverse folding problem, which is the task of designing an amino acid sequence that will fold into a specified backbone structure, represents a fundamental challenge in de novo protein design. Existing computational methods, including deep learning-based approaches, often fail to simultaneously optimize accuracy, stability, efficiency, and generalizability across diverse folds. Here, we present ProtSeqGen, a deep learning model that overcomes these limitations through a multi-stage graph-based framework. ProtSeqGen encodes protein structures as local geometric graphs, explicitly models residue-level interactions using a message-passing neural network, and predicts optimal amino acids with a multi-layer perceptron. When trained on CATH 4.2 dataset and evaluated on standard and challenging benchmarks, ProtSeqGen achieved superior sequence recovery compared to numerous state-of-the-art (SOTA) methods. It also generated accurate, designable sequences for nine topologically diverse proteins, demonstrating remarkable generalization capability. These results establish ProtSeqGen as a robust and scalable solution to the protein inverse folding problem, propelling de novo protein design with high structural precision.

BACKGROUND: The circadian clock is an evolutionarily conserved system that orchestrates 24-h physiological rhythms through transcriptional and translational feedback loops. Mounting evidence suggests a bidirectional relationship between circadian rhythm alteration and disease progression, positioning the circadian clock as a potential therapeutic target. Due to the scarcity of high-resolution temporal omics data, it remains very challenging to elucidate the underlying regulatory mechanisms of the circadian system. As a practical alternative, public untimed transcriptomic datasets offer the potential to infer gene expression oscillations retrospectively. However, existing computational approaches for circadian phase estimation often suffer from limited predictive accuracy, reducing their ability to reliably reconstruct rhythmic gene expression patterns. RESULTS: To overcome these limitations, we develop DCPR, an unsupervised deep learning framework designed to accurately reconstruct the circadian phase from untimed transcriptomic data. Through comprehensive analyses of both simulated and real data, DCPR consistently overperforms existing methods in circadian phase estimation. Additional validations using knowledgebase mining and ex vivo experimental data further support DCPR's efficacy in reconstructing the oscillatory pattern of gene expression and detecting circadian variation. CONCLUSIONS: Our study demonstrates that DCPR is a highly versatile tool for systematically identifying transcriptional rhythms from untimed expression data. This tool will facilitate therapeutics discovery for circadian-related behavioral and pathological disorders.

Brain tumor segmentation from multi-modal magnetic resonance imaging (MRI) is crucial for computer-assisted diagnosis and treatment planning. However, this task remains highly challenging due to substantial image heterogeneity, modality-inherent variability, and severe class imbalance among tumor sub-regions. To address these issues, we propose FRMF-Net , a F eature R ectification and adaptive M odality F usion guided multi-modal brain tumor segmentation Net work, which consists of three key components: a Modality-Specific Feature Rectification (MSFR) module, an Adaptive Modality Fusion (AMF) module, and a Region-Adaptive Loss (RAL). Specifically, MSFR enhances modality-specific representations by jointly modeling shared and private information, thereby mitigating inter-modality noise and reducing feature discrepancies across modalities. Building on this, AMF performs voxel-wise adaptive fusion through modality-, channel-, and spatial-wise attention, enabling the network to dynamically emphasize the most informative features for accurate tumor delineation. In addition, RAL alleviates the class imbalance issue by adaptively reweighting the contribution of each tumor sub-region according to its spatial extent in each sample. Extensive experiments on the BraTS 2019 and BraTS 2020 datasets demonstrate that FRMF-Net consistently outperforms the state-of-the-art methods, achieving superior Dice score and lower Hausdorff distance, particularly in small and challenging tumor regions. These results confirm that FRMF-Net provides a robust and effective solution for multi-modal brain tumor segmentation.

Accurate brain tumor segmentation plays an important role in clinical diagnosis, treatment planning, and therapeutic response monitoring. Multi-modal MRI provides complementary structural and functional information, but existing methods remain limited by their inadequate exploitation of cross-modal complementarity and their inability to effectively handle modality-specific disparities and redundant information. To address these challenges, this paper proposes a novel hierarchical teacher-student learning framework with adaptive cross-modal fusion. MRI modalities are grouped into teacher modalities (Flair and T1c) and student modalities (T2 and T1) based on their intrinsic tumor-related characteristics. Central to this framework is the Modality Guidance Module (MGM), which consists of two key components designed to achieve multi-modal feature distillation. Within MGM, the Modality Enhancement Module (MEM) extracts highly discriminative features from teacher modalities. While the Modality Fusion Module (MFM) leverages these features to guide and refine the learning of student modalities. To further capture inter-modal dependencies, a Cross-Modal Fusion Module (CMFM) is introduced to adaptively integrate complementary information across all modalities. Extensive experiments on the BraTS 2018, 2019 and 2020 datasets demonstrate that the proposed method achieves superior performance compared with state-of-the-art approaches. Beyond brain tumor segmentation, the hierarchical teacher-student paradigm and adaptive fusion strategy also hold potential for broader multi-modal image analysis tasks.

Accurate segmentation of brain tumors from multi-modal MRI is crucial for diagnosis and treatment planning. However, challenges such as severe class imbalance, modality-specific feature heterogeneity, and predictive uncertainty hinder reliable performance. In this work, we propose UTriGate-Net, a novel uncertainty-aware multi-modal brain tumor segmentation framework. First, we design a Triaxial Context Encoding (TCE) block that extracts anisotropic spatial features by applying directional convolutions along the axial, coronal, and sagittal planes, thereby enhancing 3D contextual representation. Second, we introduce a Gated Modality Fusion (GMF) module, which adaptively integrates complementary information across modalities through modality-specific gating weights that suppress redundancy while retaining salient features. Finally, to improve segmentation reliability, we develop an Uncertainty-Regularized Weighted Loss (URWL) that combines dynamic class-specific weighting to mitigate class imbalance with an entropy-based uncertainty penalty to encourage well-calibrated predictions. Experiments on the BraTS 2019 and 2020 datasets demonstrate that UTriGate-Net achieves superior segmentation accuracy and robustness, particularly in challenging subregions. Overall, the proposed framework offers a promising solution for reliable and precise brain tumor delineation in clinical practice.

Accurate brain tumor segmentation from multimodal MRI is critical for clinical diagnosis and treatment planning. However, effectively leveraging the complementary information across different modalities remains a significant challenge due to modality-specific noise, information redundancy and inherent model uncertainty. To tackle these challenges, we propose a Disentangled Fusion Network (DFuse-Net) that integrates disentangled feature fusion with contrastive and consistency-aware learning to enable reliable multi-modal brain tumor segmentation. Our method first explicitly disentangles modality-shared and modality-specific feature representations. Then, a Disentangled Feature Fusion Module (DFFM) is proposed to effectively integrate modality-shared and modalityspecific feature representations. In addition, a contrastive-aware learning scheme is employed to enhance feature discriminability, while a consistency-aware learning strategy is applied to enforce structural coherence across modalities. Moreover, Monte Carlo dropout is applied during inference to generate voxelwise aleatoric and epistemic uncertainty maps, enhancing the robustness of segmentation. Extensive experiments on the BraTS datasets demonstrate that DFuse-Net achieves superior segmentation accuracy and reliability compared to the state-of-the-art methods.

BACKGROUND: The exploration of drug-target interactions (DTIs) is a critical step in drug discovery and drug repurposing. Recently, network-based methods have emerged as a prominent research area for predicting DTIs. These methods excel by extracting both topological and feature information from DTIs networks, thereby achieving superior DTIs prediction performance. However, the majority of existing GCN-based methods utilize shallow graph neural networks, which are incapable of extracting higher-level semantic information. Additionally, the current training of models lacks an effective guiding mechanism, leading to the insufficient improvement of network's representation capabilities. RESULTS: In this paper, we propose a graph convolutional autoencoder model, named DDGAE, for DTIs prediction. We develop a DWR-GCN module, which incorporates dynamic weighting graph convolution with residual connection, to improve the representation capability for DTI heterogeneous networks. Further, to improve the learning efficiency of the model, we devise a dual self-supervised joint training mechanism. Specifically, this mechanism integrates DWR-GCN and a graph convolutional autoencoder into a cohesive system, enhancing both the learning performance and stability of DDGAE. CONCLUSION: Experimental results show that DDGAE significantly outperforms several SOTA models in DTIs prediction, achieving optimal performance and the reliability of our method is verified by case study.

Efficient and accurate recognition of protein-DNA interactions is vital for understanding the molecular mechanisms of related biological processes and further guiding drug discovery. Although the current experimental protocols are the most precise way to determine protein-DNA binding sites, they tend to be labor-intensive and time-consuming. There is an immediate need to design efficient computational approaches for predicting DNA-binding sites. Here, we proposed ULDNA, a new deep-learning model, to deduce DNA-binding sites from protein sequences. This model leverages an LSTM-attention architecture, embedded with three unsupervised language models that are pre-trained on large-scale sequences from multiple database sources. To prove its effectiveness, ULDNA was tested on 229 protein chains with experimental annotation of DNA-binding sites. Results from computational experiments revealed that ULDNA significantly improves the accuracy of DNA-binding site prediction in comparison with 17 state-of-the-art methods. In-depth data analyses showed that the major strength of ULDNA stems from employing three transformer language models. Specifically, these language models capture complementary feature embeddings with evolution diversity, in which the complex DNA-binding patterns are buried. Meanwhile, the specially crafted LSTM-attention network effectively decodes evolution diversity-based embeddings as DNA-binding results at the residue level. Our findings demonstrated a new pipeline for predicting DNA-binding sites on a large scale with high accuracy from protein sequence alone.

Most life activities in organisms are regulated through protein complexes, which are mainly controlled via Protein-Protein Interactions (PPIs). Discovering new interactions between proteins and revealing their biological functions are of great significance for understanding the molecular mechanisms of biological processes and identifying the potential targets in drug discovery. Current experimental methods only capture stable protein interactions, which lead to limited coverage. In addition, expensive cost and time consuming are also the obvious shortcomings. In recent years, various computational methods have been successfully developed for predicting PPIs based only on protein homology, primary sequences of protein or gene ontology information. Computational efficiency and data complexity are still the main bottlenecks for the algorithm generalization. In this study, we proposed a novel computational framework, HNSPPI, to predict PPIs. As a hybrid supervised learning model, HNSPPI comprehensively characterizes the intrinsic relationship between two proteins by integrating amino acid sequence information and connection properties of PPI network. The experimental results show that HNSPPI works very well on six benchmark datasets. Moreover, the comparison analysis proved that our model significantly outperforms other five existing algorithms. Finally, we used the HNSPPI model to explore the SARS-CoV-2-Human interaction system and found several potential regulations. In summary, HNSPPI is a promising model for predicting new protein interactions from known PPI data.

A deep transfer learning framework adapting mixed subdomains is proposed for cross-species plant disease diagnosis. Most existing deep transfer learning studies focus on knowledge transfer between highly correlated domains. These methods may fail to deal with domains that are poorly correlated. In this study, mixed domain images were generated from source and target image groups for improving the correlation between the mixed domain (training dataset) and the target domain (testing dataset). A subdomain alignment mechanism is employed to transfer knowledge from the mixed domain to the target domain. The proposed framework captures the fine-grained information more effectively. Extensive experiments were conducted and prove that the proposed method produces a more effective result compared with existing deep transfer learning technologies for poorly related subdomains.

Pathogenicity-related studies are of great importance in understanding the pathogenesis of complex diseases and improving the level of clinical medicine. This work proposed a bioinformatics scheme to analyze cancer-related gene mutations, and try to figure out potential genes associated with diseases from the protein domain-domain interaction network. Herein, five measures of the principle of centrality lethality had been adopted to implement potential correlation analysis, and prioritize the significance of genes. This method was further applied to KEGG pathway analysis by taking the malignant melanoma as an example. The experimental results show that 25 domains can be found, and 18 of them have high potential to be pathogenically important related to malignant melanoma. Finally, a web-based tool, named Human Cancer Related Domain Interaction Network Analyzer, is developed for potential pathogenic genes prioritization for 26 types of human cancers, and the analysis results can be visualized and downloaded online.

The papers in this special section focus on machine learning for artificial intelligent-enhances healthcare and medical services. These services are always among the top concerns for humans, especially under the special situation of COVID-19 pandemic, started from early 2020. In the field of computational biology and bioinformatics, scientists seek various possibilities using computer technologies, especially artificial intelligence (AI) enhanced methods, for healthcare services and medical diagnoses. For example, over the past few years, scientists have been working hard to identify the internal relationships between gene microarrays, cells, tissues, organisms, diseases, etc., and apply the AI, machine learning and deep learning technologies looking for more innovative solutions for new diseases, such as COVID-19. In fact, nowadays, AI technology, such as the convolutional neural network (CNN), is considered has one of the most important computer technologies and has been widely applied in the fields of healthcare engineering, medical research, disease diagnosis, cancer/tumor analysis and etc.

Artificial intelligence-based air quality index (AQI) forecasting is a hot research topic in the fields of sustainable and smart industrial environment design. There are mainly two obstacles that hinder the existing machine learning (ML) and deep learning (DL) technologies providing accurate forecasting results to protect the environment, which include the intercorrelation between different AQI components and the highly volatile AQI pattern changes. In this article, a novel DL framework combining multiple nested long short term memory networks (MTMC-NLSTM) is proposed for accurate AQI forecasting enlightened with the federated learning. The performance of the proposed MTMC-NLSTM model is compared with conventional ML models, DL methods, as well as hybrid DL models. The experimental results show that the performance of the proposed method is superior to those of all compared models.

Cancer cell detection and its stages recognition of life cycle are an important step to analyze cellular dynamics in the automation of cell based-experiments. In this work, a two-stage hierarchical method is proposed to detect and recognize different life stages of bladder cells by using two cascade Convolutional Neural Networks (CNNs). Initially, a hybrid object proposal algorithm (called EdgeSelective) by combining EdgeBoxes and Selective Search is proposed to generate candidate object proposals instead of a single Selective Search method in Region-CNN (R-CNN), and it can exploit the advantages of different mechanisms for generating proposals so that each cell in the image can be fully contained by at least one proposed region during the detection process. Then, the obtained cells from the previous step are used to train and extract features by employing CNNs for the purpose of cell life stage recognition. Finally, a series of comparison experiments are implemented. The results show that the proposed method can obtain better performance than traditional methods either in the stage of cell detection or cell life stage recognition, and it encourages and suggests the application in the development of new anticancer drug and cytopathology analysis of cancer patients in the near future.

Discord search is an important technique for time series analysis, especially for anomaly detection . In recent years, many computational approaches of discord search were studied; however, limitation exists while only the problems with univariate time series data can be well addressed. In this study, we proposed a novel computational framework to identify discords from multivariate time series (MTS) data, namely, LRRDS (Local Recurrence Rate based Discord Search). LRRDS accurately identifies the discords by analyzing a recurrence plot, which is transformed from the original time series data. An innovative strategy was employed to improve the efficiency for pair-wise distance comparison of two subsequences . In the experimental simulations, LRRDS was applied to an extensive number of MTS datasets. Results show that the proposed approach is more efficient than existing methods, such as GDS. In conclusion, the LRRDS approach solves the adaptability problem of discord sequences in multi-dimensional space and guarantees the computational effectiveness and efficiency.

Prostate cancer (PCa) is the most commonly diagnosed malignancy and the second leading cause of cancer-related death in American men. Androgen deprivation therapy (ADT) has become a standard treatment strategy for advanced PCa. Although a majority of patients initially respond to ADT well, most of them will eventually develop castration-resistant PCa (CRPC). Previous studies suggest that ADT-induced changes in the immune microenvironment (mE) in PCa might be responsible for the failures of various therapies. However, the role of the immune system in CRPC development remains unclear. To systematically understand the immunity leading to CRPC progression and predict the optimal treatment strategy in silico, we developed a 3D Hybrid Multi-scale Model (HMSM), consisting of an ODE system and an agent-based model (ABM), to manipulate the tumor growth in a defined immune system. Based on our analysis, we revealed that the key factors (e.g. WNT5A, TRAIL, CSF1, etc.) mediated the activation of PC-Treg and PC-TAM interaction pathways, which induced the immunosuppression during CRPC progression. Our HMSM model also provided an optimal therapeutic strategy for improving the outcomes of PCa treatment.

The extreme learning machine (ELM) is famous for its single hidden-layer feed-forward neural network which results in much faster learning speed comparing with traditional machine learning techniques. Moreover, extensions of ELM achieve stable classification performances for imbalanced data. In this paper, we introduce a hybrid method combining the extended Kalman filter (EKF) with cost-sensitive dissimilar ELM (CS-D-ELM). The raw data are preprocessed by EKF to produce inputs for the CS-D-ELM classifier. Experimental results show that the proposed method is more suitable for real-time fault diagnosis of air handling units than traditional approaches.

BACKGROUND: Ion mobility-mass spectrometry (IMMS), an analytical technique which combines the features of ion mobility spectrometry (IMS) and mass spectrometry (MS), can rapidly separates ions on a millisecond time-scale. IMMS becomes a powerful tool to analyzing complex mixtures, especially for the analysis of peptides in proteomics. The high-throughput nature of this technique provides a challenge for the identification of peptides in complex biological samples. As an important parameter, peptide drift time can be used for enhancing downstream data analysis in IMMS-based proteomics. RESULTS: In this paper, a model is presented based on least square support vectors regression (LS-SVR) method to predict peptide ion drift time in IMMS from the sequence-based features of peptide. Four descriptors were extracted from peptide sequence to represent peptide ions by a 34-component vector. The parameters of LS-SVR were selected by a grid searching strategy, and a 10-fold cross-validation approach was employed for the model training and testing. Our proposed method was tested on three datasets with different charge states. The high prediction performance achieve demonstrate the effectiveness and efficiency of the prediction model. CONCLUSIONS: Our proposed LS-SVR model can predict peptide drift time from sequence information in relative high prediction accuracy by a test on a dataset of 595 peptides. This work can enhance the confidence of protein identification by combining with current protein searching techniques.