Subtomogram alignment is a critical task in cryo-electron tomography (cryo-ET) analysis, essential for achieving high-resolution reconstructions of macromolecular complexes. However, learning effective positional representations remains challenging due to limited labels and high noise levels inherent in cryo-ET data. In this work, we address this challenge by proposing a self-supervised learning approach that leverages intrinsic geometric transformations as implicit supervisory signals, enabling robust representation learning despite data scarcity. We introduce BOE-ViT, the first Vision Transformer (ViT) framework for 3D subtomogram alignment. Recognizing that traditional ViTs lack equivariance and are therefore suboptimal for orientation estimation, we enhance the model with two innovative modules that introduce equivariance include 1) the Polyshift module for improved shift estimation and 2) Multi-Axis Rotation Encoding (MARE) for enhanced rotation estimation. Experimental results demonstrate that BOE-ViT significantly outperforms state-of-the-art methods. Notably, at SNR 0.01 dataset, our approach achieves a 77.3% reduction in rotation estimation error and a 62.5% reduction in translation estimation error, effectively overcoming the challenges in cryo-ET subtomogram alignment.

Skin lesion segmentation is crucial in computer aided diagnosis and treatment of skin diseases. Lesions with varying sizes, blurred boundaries, and irregular shapes are challenging to segment. Both CNNs-based and transformer-based models have been extensively explored in medical image segmentation. However, CNNs have limitations in capturing long-range dependencies. Transformers are hindered by their quadratic computational complexity. Recently, Mamba, as a state-space model (SSM), has shown promise in modeling long-range interactions while maintaining linear computational complexity. To address these issues, we propose an effective multi-scale adaptive enhancement network for skin disease image segmentation, named MAE2-Net. Specifically, we introduce a multi-scale feature adaptive module(MSFA), which can promote feature fusion at skip connections by dynamically weighting and combining coarse-grained encoder features with fine-grained decoder features. We also incorporate a grouped parallel feature enhancement module(GPFE), which can leverage the advantages of Mamba and xLSTM to enhance feature representations while reducing computational burden through parallel processing. Additionally, we propose a dynamic visual KAN module(DVKM), which can efficiently extract multi-scale features and utilizes the interpretability of Kolmogorov-Arnold Network(KAN) to deeply transform and combine features, thereby extracting more advanced feature representations. Our proposed MAE2-Net has been evaluated on two public skin lesion segmentation datasets, and the experimental results demonstrate that it achieves more accurate segmentation performance, making it more competitive compared to existing SOTA methods.

Illuminating interactions between proteins and small drug molecules is a long-standing challenge in the field of drug discovery. Despite the importance of understanding these interactions, most previous works are limited by hand-designed scoring functions and insufficient conformation sampling. The recently-proposed graph neural network-based methods provides alternatives to predict protein-ligand complex conformation in a one-shot manner. However, these methods neglect the geometric constraints of the complex structure and weaken the role of local functional regions. As a result, they might produce unreasonable conformations for challenging targets and generalize poorly to novel proteins. In this paper, we propose Trigonometry-Aware Neural networKs for binding structure prediction, TANKBind, that builds trigonometry constraint as a vigorous inductive bias into the model and explicitly attends to all possible binding sites for each protein by segmenting the whole protein into functional blocks. We construct novel contrastive losses with local region negative sampling to jointly optimize the binding interaction and affinity. Extensive experiments show substantial performance gains in comparison to state-of-the-art physics-based and deep learning-based methods on commonly-used benchmark datasets for both binding structure and affinity predictions with variant settings.