Non-small-cell lung cancer (NSCLC) represents most of lung cancers, is often diagnosed at an advanced metastatic stage. Therefore, exploring the mechanisms underlying metastasis is key to understanding the development of NSCLC. The expression of B cell receptor-associated protein 31 (BCAP31), calreticulin, glucose-regulated protein 78, and glucose-regulated protein 94 were analyzed using immunohistochemical staining of 360 NSCLC patients. It resulted that the high-level expression of the four proteins, but particularly BCAP31, predicted inferior overall survival. What's more, BCAP31 was closely associated with histological grade and p53 status, which was verified by seven cohorts of NSCLC transcript microarray datasets. Then, three NSCLC cell lines were transfected to observe behavior changes BCAP31 caused, we found the fluctuation of BCAP31 significantly influenced the migration, invasion of NSCLC cells. To identify the pathway utilized by BCAP31, Gene Set Enrichment Analysis was firstly performed, showing Akt/m-TOR/p70S6K pathway was the significant one, which was verified by immunofluorescence, kinase phosphorylation and cellular behavioral observations. Finally, the data of label-free mass spectroscopy implied that BCAP31 plays a role in a fundamental biological process. This study provides the first demonstration of BCAP31 as a novel prognostic factor related to metastasis and suggests a new therapeutic strategy for NSCLC.Lung cancer is one of the most prevalent neoplasms and the leading cause of cancer-related death worldwide, being responsible for nearly one in five cases 1 . Non-small-cell lung cancer (NSCLC) contributes to 85% of lung cancer cases. Owing to the absence of clinical symptoms and effective screening programs, most lung cancers are diagnosed at an advanced stage with metastasis. Targeted immunotherapy, including anti-angiogenic and checkpoint monoclonal antibodies or tyrosine kinase inhibitors, demonstrates better efficacy than traditional surgical treatment and radio-chemotherapy, although drug resistance and tumor heterogeneity remain significant problems, and metastasis remains the major cause of mortality 2 .It is therefore important to identify efficient symbolic markers of metastasis and therapeutic targets for NSCLC. Given the aberrant expression of specific genes in a variety of cancer types, restricted in testis or selected in normal tissue, cancer-testis antigens (CTAs) have emerged as efficient specific tumor targets which spare normal tissue from incurring damage during treatment 3 . Originally described in patients with malignant melanoma 4 , CTAs have been identified as biomarkers for a diverse range of cancers, including NSCLC 5 . Their expression is often coordinated 6 , and associated with poor clinical outcome 7 and advanced stage 8 , particularly metastasis 9 .with histological grade (p = 0.0009) and p53 status (p = 0.0058; Supplementary Table S1). However, there was no correlation between BCAP31 mRNA expression and NSCLC prognosis, which was inconsistent with our protein ...

No abstract

No abstract

A local distance comparison for multiple-shot people re-identification based on a new adaptive metric learning method is introduced in this paper. There exist two intrinsic issues in multiple-shot person re-identification: Large variances in view point, illumination, and non-rigid deformation are included in the image set of the same person; only a few training data for learning tasks are available in a realistic reidentification scenario. We deal with the multimodal property of people's appearance distribution caused by the first issue by using a local distance comparison approach. Since the capability of the local distance comparison highly depends on the choice of distance metric, we also introduce an adaptive learning method to learn an appropriate distance metric and use it to find and compute local neighbors effectively. This adaptive learning method is able to solve the overfitting problem caused by the second issue, through leveraging the generic knowledge of re-identification together with the specific information of the target task. We evaluated our approach on public benchmark datasets, and confirmed its superiority as compared to conventional approaches.

No abstract

In this paper, a deep learning method for video-based action recognition is proposed. On the one hand, boundary compensation on the basis of a deep neural network is performed to achieve action proposal. Boundary compensation considering non-maximum suppression according to sliding window priority is applied to remove redundant windows. To accurately detect boundaries, a boundary compensation network is established with multiple networks to process different numbers of segments. On the other hand, action recognition based on the resultant action proposals is performed. To further utilise boundary compensation, three methods are introduced for key frame selection. Optical flow and RGB features are combined via a channel fusion to realise feature representation. A two-stream network with a spatiotemporal structure is adopted for action recognition. The proposed method is evaluated on three public datasets. The experimental results demonstrate that the proposed method achieves a superior performance to that of state-of-the-art methods.