Search

Background: Degrons are short linear motifs, bound by E3 ubiquitin ligase to target protein substrates to be degraded by the ubiquitin-proteasome system. Mutations leading to deregulation of degron functionality disrupt control of protein abundance due to mistargeting of proteins destined for degradation and often result in pathologies. Targeting degrons by small molecules also emerges as an exciting drug design strategy to upregulate the expression of specific proteins. Despite their essential function and disease targetability, reliable identification of degrons remains a conundrum. Here, we developed a deep learning-based model named Degpred that predicts general degrons directly from protein sequences., Results: We showed that the BERT-based model performed well in predicting degrons singly from protein sequences. Then, we used the deep learning model Degpred to predict degrons proteome-widely. Degpred successfully captured typical degron-related sequence properties and predicted degrons beyond those from motif-based methods which use a handful of E3 motifs to match possible degrons. Furthermore, we calculated E3 motifs using predicted degrons on the substrates in our collected E3-substrate interaction dataset and constructed a regulatory network of protein degradation by assigning predicted degrons to specific E3s with calculated motifs. Critically, we experimentally verified that a predicted SPOP binding degron on CBX6 prompts CBX6 degradation and mediates the interaction with SPOP. We also showed that the protein degradation regulatory system is important in tumorigenesis by surveying degron-related mutations in TCGA., Conclusions: Degpred provides an efficient tool to proteome-wide prediction of degrons and binding E3s singly from protein sequences. Degpred successfully captures typical degron-related sequence properties and predicts degrons beyond those from previously used motif-based methods, thus greatly expanding the degron landscape, which should advance the understanding of protein degradation, and allow exploration of uncharacterized alterations of proteins in diseases. To make it easier for readers to access collected and predicted datasets, we integrated these data into the website http://degron.phasep.pro/ ., (© 2022. The Author(s).)

With people's higher and higher spiritual requirements, fiber art is more common in people's lives. From ordinary fiber materials to handicrafts and daily necessities, there are shadows of fiber art. This paper aims to study the application of two-dimensional images in the visual and tactile dimensions of fiber art design. This paper proposes a three-dimensional simulation of two-dimensional images of art based on deep learning technology to achieve the purpose of convenient display and also analyzes the feature extraction in fiber art. In order to comprehensively compare the performance of the algorithms, this paper compares the performance of four categories: reconstruction speed, number of reconstructed point clouds, diversity of reconstructable categories, and reconstruction stability. The experimental results show that the number of point clouds obtained by the algorithm proposed in this paper far exceeds the other two reconstruction processes, and the average time is 34.4 min. This also shows that such algorithmic processes are more robust to clothing diversity. [ABSTRACT FROM AUTHOR]

Person re-identification aims to retrieve specific pedestrians from different cameras and scenes, in which extracting robust and discriminative features is crucial for this task. To explore the potential interactions among images and learn more robust representations, this paper proposes Transformer-based Feature Interactor(TFI) and improved Margin Self-punishment Softmax loss(MS-Softamx). The Transformer-based Feature Interactor consists of Group Channel Pyramid Attention(GCPA) and Neighbor Interaction Modeling(NIM). Firstly, the Group Channel Pyramid Attention module provides prior information for high-level semantics via low-level semantics. The attention information is gradually stacked from coarse to fine to obtain enhanced hierarchical multi-scale features. Then, Neighbor Interaction Modeling effectively model the input and similar neighbors to produce a more robust and discriminative image representation. To make TFI more focused on intra-class embedding learning, we also propose Margin Self-punishment Softmax guide deep network learning, which obtains a tighter custom classification boundary by pushing the inter-class threshold and minimizing the intra-class variance. The proposed method is verified on four datasets, and this achieves 92.8%/95.6% mAP/Rank-1 on Market1501, 86.1%/90.8% mAP/Rank-1 on DukeMTMC, 64.4%/ 81.2% mAP/Rank-1 on MSMT17, 79.7%/80.8% mAP/Rank-1 on CUHK03-detected and 81.8%/81.9% mAP/Rank-1 on CUHK03-labeled. Extensive experiments demonstrate that the proposed method achieves competitive performance with other state-of-the-art methods. • A Transformer-based network for multi-scale information interaction. • Constructing context-aware features containing global information and detailed features with pyramidal attention. • Adaptive penalty margin for tighter custom classification boundary. [ABSTRACT FROM AUTHOR]