RA-Net: reverse attention for generalizing residual learning.

Since residual learning was proposed, identity mapping has been widely utilized in various neural networks. The method enables information transfer without any attenuation, which plays a significant role in training deeper networks. However, interference with unhindered transmission also affects the network's performance. Accordingly, we propose a generalized residual learning architecture called reverse attention (RA), which applies high-level semantic features to supervise low-level information in the identity mapping branch. It means that higher semantic features selectively transmit low-level information to deeper layers. In addition, we propose a Modified Global Response Normalization(M-GRN) to implement reverse attention. RA-Net is derived by embedding M-GRN in the residual learning framework. The experiments show that the RA-Net brings significant improvements over residual networks on typical computer vision tasks. For classification on ImageNet-1K, compared with resnet101, RA-Net improves the Top-1 accuracy by 1.7% with comparable parameters and computational cost. For COCO detection, on Faster R-CNN, reverse attention improves box AP by 1.9%. Meanwhile, reverse attention improves UpperNet's mIoU by 0.7% on ADE20K segmentation.

A bioinformatics framework to identify the biomarkers and potential drugs for the treatment of colorectal cancer.

Colorectal cancer (CRC), a common malignant tumor, is one of the main causes of death in cancer patients in the world. Therefore, it is critical to understand the molecular mechanism of CRC and identify its diagnostic and prognostic biomarkers. The purpose of this study is to reveal the genes involved in the development of CRC and to predict drug candidates that may help treat CRC through bioinformatics analyses. Two independent CRC gene expression datasets including The Cancer Genome Atlas (TCGA) database and GSE104836 were used in this study. Differentially expressed genes (DEGs) were analyzed separately on the two datasets, and intersected for further analyses. 249 drug candidates for CRC were identified according to the intersected DEGs and the Crowd Extracted Expression of Differential Signatures (CREEDS) database. In addition, hub genes were analyzed using Cytoscape according to the DEGs, and survival analysis results showed that one of the hub genes, TIMP1 was related to the prognosis of CRC patients. Thus, we further focused on drugs that could reverse the expression level of TIMP1. Eight potential drugs with documentary evidence and two new drugs that could reverse the expression of TIMP1 were found among the 249 drugs. In conclusion, we successfully identified potential biomarkers for CRC and achieved drug repurposing using bioinformatics methods. Further exploration is needed to understand the molecular mechanisms of these identified genes and drugs/small molecules in the occurrence, development and treatment of CRC.

Spatiotemporal of the Coupling Relationship between Ecosystem Services and Human Well-Being in Guanzhong Plain Urban Agglomeration.

Understanding the complex relationship between ecosystem services and human well-being during the rapid development of urban agglomerations can promote the sustainable development of urban agglomerations. In this paper, the InVEST model and ArcGIS10.2 were used to analyze the spatial and temporal evolution characteristics of ecosystem services and human well-being in the Guanzhong Plain urban agglomeration. On this basis, the coupling coordination index is used to reveal the spatiotemporal coupling relationship between them. (1) From 2010 to 2018, the water conservation services, soil conservation services, and carbon sequestration services of the Guanzhong Plain urban agglomeration showed a fluctuating downward trend. The spatial differences of ecosystem services were significant. (2) From 2010 to 2018, human well-being in the Guanzhong Plain urban agglomeration showed a fluctuating downward trend, with a decrease of 17%, and regional differences tended to narrow. (3) The coupling coordination degree between ecosystem services and human well-being has slightly decreased while maintaining the basic coordination state. The results show that there was a significant relationship between the decline of ecosystem services and the rapid development of the Guanzhong Plain urban agglomeration, and policies should be classified according to the coupling coordination types of human well-being and ecosystem services to promote the sustainable development of urban agglomerations.

Long noncoding RNA SNHG6 functions as a competing endogenous RNA by sponging miR-181a-5p to regulate E2F5 expression in colorectal cancer.

BACKGROUND: Long noncoding RNAs (lncRNAs) have been identified as a novel class of regulators implicated in diverse biological processes in human cancers. Currently, evidence have shown that SNHG6, a cancer-associated lncRNA, exerts critical functions in gastric cancer and hepatocellular carcinoma; however, its role in colorectal cancer (CRC) remains unclear. METHODS: The expression of SNHG6 was determined by quantitative real-time PCR in CRC tissues and cells. SNHG6 was downregulated by using RNAi technology. Cell proliferation was examined by MTT and clone formation assays. Cell migration and invasion were determined by wound healing and transwell assays. Fluorescence in situ hybridization assays were performed to examine subcellular localization of SNHG6 in CRC cells. Fluorescence reporter and Western blot assays were used to explore the potential mechanisms of SNHG6 in CRC progression. RESULTS: In this study, we found that SNHG6 was significantly upregulated in CRC tissues and cell lines, compared with normal tissues and normal colorectal epithelial cell line NCM460, respectively. High expression of SNHG6 was positively correlated with tumor size, advanced TNM stage, and distant metastasis. Survival analyses revealed that SHNG6 was significantly associated with poor clinical outcomes and could serve as an independent prognostic factor. Loss-of-function studies demonstrated that SNHG6 knockdown inhibited CRC cell proliferation, induced G0/G1 arrest, promoted apoptosis, suppressed CRC cell migration and invasion, and restrained tumor growth. Mechanistic investigations showed that SNHG6 acted as a competing endogenous RNA for miR-181a-5p and attenuated the inhibitory effect of miR-181a-5p on E2F5. CONCLUSION: Taken together, these results demonstrated that SNHG6 plays a crucial role in CRC progression via miR-181a-5p/E2F5 axis. Therefore, SNHG6 may serve as a prognostic and therapeutic biomarker in CRC.