Drug-target interaction (DTI) prediction plays a pivotal role in accelerating drug discovery. Nevertheless, existing AI-driven approaches face three critical limitations: existing attention-based methods lack dynamic bidirectional interaction channels, limiting their ability to model asymmetric drug-target communication patterns; conventional architectures struggle to integrate both local binding patterns and global biological contexts; and rigid prediction heads discarding spatial interaction patterns. These issues hamper the accuracy and comprehensiveness of DTI prediction. To address these, we propose CAMA-DTI, an end-to-end framework integrating three innovations. First, a cross-domain bidirectional attention module establishes dual-perspective interaction channels that enable co-evolutionary feature refinement through mutual pharmacological feedback. Second, the Mamba-driven fusion block incorporates state-space modeling to dynamically integrate local binding patterns with global biological contexts across extended sequences. Third, we replace conventional classifiers with Kolmogorov-Arnold Networks (KANs) employing adaptive spline transformations that preserve multi-scale interaction signatures while maintaining parametric efficiency. Extensive experimental results demonstrate that CAMA-DTI achieves robust and accurate predictions across diverse datasets, outperforming state-of-the-art methods in both established and novel drug target scenarios. Notably, the framework maintains consistent performance across datasets of varying scales, and case studies validate its practical utility in real-world drug development pipelines.

Functional magnetic resonance imaging (fMRI) allows the observation of brain functional connectivity patterns. Attention-based diagnostic models have been widely applied in fMRI data for brain disease diagnosis. However, the global attention mechanism of the Transformer faces challenges in adaptively identifying and focusing on significant brain regions and connections relevant to disease diagnosis while reducing attention to non-relevant regions and connections in fMRI data, as well as the degradation problem of the attention mechanism, thereby limiting the improvement in diagnostic accuracy. To address these problems, we propose a connection-mask-residual focused attention network (Trifocal Transformer) based on fMRI data for brain disease diagnosis. In the Trifocal Transformer, a Connection Focus Module is developed to simulate brain functional connectivity, thereby enhancing the attention mechanism's ability to focus on significant regions and connections relevant to disease diagnosis. To mitigate the potential negative impact of non-focused regions in the attention map, a learnable Mask Focus Module is designed to adaptively reduce attention to non-relevant regions and connections. To address the degradation of the attention mechanism's focusing ability, we establish Residual Focus Connections between the attention maps, which reinforce the focusing effect across layers and ensure stable attention to significant features. Comprehensive experimental results demonstrate that the Trifocal Transformer achieves superior diagnostic accuracies of 74.1% and 71.2% on ADHD-200 and ABIDE I datasets, respectively. Furthermore, our method reveals potentially disease-related regions of interest (ROIs), providing a new neuroimaging perspective for brain disease diagnosis and treatment.

Inferring the proportion of cell types present in each spatial location from a mixture of expression profiles in spatial transcriptome (ST) data is a crucial task. However, on the one hand, existing methods adopt static or predefined neighborhood structures for ST data, which limits their ability to adapt to complex spatial organization. On the other hand, existing models often rely on overly simplified linear hybrid models and fail to effectively capture spatial dependencies. To address these limitations, we propose a new deconvolution framework called NSAMTG. Specifically, we utilize the Native Sparse Attention transformer with kernel mean matching (KMM) to reduce the distribution difference of the data and perform encoding reconstruction on single-cell RNA sequencing (scRNA-seq) data. At the same time, we build a Mixed-Topology graph transformer VAE that combines an adaptive adjacency matrix and a static adjacency matrix to capture spatial expression correlations and spatial relationships in ST data. Finally, we employ a Generative Adversarial Network with a FourierKAN discriminator to learn the complex, non-linear cell-to-spot mapping relationship and achieve cell type deconvolution. Experiments on both simulated datasets and real datasets show that our method is superior to the existing methods, opening up new possibilities for robust and zeneralizable spatial deconvolution in diverse biological contexts.

Accurate simulation of the brain's intrinsic dynamic activity is essential for understanding human cognition and behavior and developing personalized brain disease therapies. Traditional neurodynamics models depend on structural connectivity to explain the emergence of functional connectivity (FC). However, achieving high-fidelity simulations at the individual level remains challenging, as the models fail to fully capture the brain information. To address these challenges, we introduce the Stable Individualized Brain Computing Model (SI-BCM), a data-driven reverse engineering framework designed to infer spatiotemporal co-activity patterns from fMRI data for simulating whole-brain activity. This model captures the dynamic interactions between brain regions by integrating spatiotemporal dimensional information to extract a stable and shared connectivity pattern, representing the intrinsic functional collaboration pattern of the brain. This connectivity pattern is then used as the core connection weight in the dynamical system. Additionally, the model has a new cost function based on the Phase-Space Association matrix (PSA), enhancing its ability to capture brain activity dynamics. This combination enables the SI-BCM to improve simulation accuracy at the individual level compared to existing models, achieving a correlation coefficient between simulated and empirical FC of 0.87. The SI-BCM also showed enhanced robustness and reliability, and effectively captured brain properties. We found the model sensitively reflected changes in cognitive function, thereby providing valuable insights into the underlying neural mechanisms. Furthermore, the application of SI-BCM in the brain modeling of Alzheimer's disease (AD) patients substantiated the hypothesis that AD pathogenesis may be due to excessive neuronal excitation. This work establishes a new paradigm for brain network modeling by prioritizing the inference of stable dynamics features from activity data, providing a powerful tool for understanding brain function and pathophysiology.

The specific binding between antibodies (Ab) and antigens (Ag) is crucial for developing drugs and vaccines to treat major diseases. Therefore, accurate identification of antibody-antigen interactions (AAI) is crucial for a comprehensive understanding of antibody therapeutic mechanisms. While wet-lab methods accurately characterize AAI, they require significant human, financial, and time costs. Traditional computational methods help to reduce the resource consumption of AAI identification, but suffer from several problems, such as (1) they rely solely on sequence data, ignoring critical 3D structural determinants; (2) the scarcity of data on antibody-antigen interactions severely limits existing methods' ability to represent unseen antibodies; (3) they focus narrowly on paratope-epitope residues, overlooking the contextual information provided by distal non-binding regions that can influence interaction patterns. To address these issues, we present an innovative model that learns sequence and structural representations to predict antibody-antigen interactions (SSRAAI). We extracted structural features by constructing contact maps from predicted PDB 3D structures. Additionally, the integration of sequence features based on adaptive relational graphs led to enhanced prediction outcomes. Our approach offers a unique integration of 3D structural information from PDB with sequence data, applied directly to Ab and Ag. Comparative results on two datasets, HIV and SARS-CoV-2, demonstrate the validity of our approach in identifying AAIs.