ABSTRACT
Background: FRAS1 (Fraser syndrome protein 1), together with FREM1 (the Fras1-related extracellular matrix proteins 1) and FREM2, belonging to the FRAS1/FREM extracellular matrix protein family, are considered to play essential roles in renal organogenesis and cancer progression. However, their roles in kidney renal clear cell carcinoma (KIRC) remain to be elucidated. Methods: FRAS1/FREM RNA expression analysis was performed using TCGA/GTEx databases, and valided using GEO databases and real-time PCR. Protein expression was peformed using CPTAC databases. Herein, we employed an array of bioinformatics methods and online databases to explore the potential oncogenic roles of FRAS1/FREM in KIRC. Results: We found that FRAS1, FREM1 and FREM2 genes and proteins expression levels were significantly decreased in KIRC tissues than in normal tissues. Decreased FRAS1/FREM expression levels were significantly associated with advanced clinicopathological parameters (pathological stage, grade and tumor metastasis status). Notably, the patients with decreased FRAS1/FREM2 expression showed a high propensity for metastasis and poor prognosis. FRAS1/FREM were correlated with various immune infiltrating cells, especially CD4+ T cells and its corresponding subsets (Th1, Th2, Tfh and Tregs). FRAS1 and FREM2 had association with DNA methylation and their single CpG methylation levels were associated with prognosis. Moreover, FRAS1/FREM might exert antitumor effects by functioning in key oncogenic signalling pathways and metabolic pathways. Drug sensitivity analysis indicated that high FRAS1 and FREM2 expression can be a reliable predictor of targeted therapeutic drug response, highlighting the potential as anticancer drug targets. Conclusion: Together, our results indicated that FRAS1/FREM family members could be potential therapeutic targets and valuable prognostic biomarkers of KIRC.
ABSTRACT
Deep neural networks are fragile under adversarial attacks. In this work, we propose to develop a new defense method based on image restoration to remove adversarial attack noise. Using the gradient information back-propagated over the network to the input image, we identify high-sensitivity keypoints which have significant contributions to the image classification performance. We then partition the image pixels into the two groups: high-sensitivity and low-sensitivity points. For low-sensitivity pixels, we use a total variation (TV) norm-based image smoothing method to remove adversarial attack noise. For those high-sensitivity keypoints, we develop a structure-preserving low-rank image completion method. Based on matrix analysis and optimization, we derive an iterative solution for this optimization problem. Our extensive experimental results on the CIFAR-10, SVHN, and Tiny-ImageNet datasets have demonstrated that our method significantly outperforms other defense methods which are based on image de-noising or restoration, especially under powerful adversarial attacks.
