Circular RNAs (circRNAs) have been regarded as critical regulators of human diseases and biological markers in some types of malignancies, including pancreatic ductal adenocarcinoma (PDAC). Recently, circ_0007534 has been identified as a novel cancer-related circRNA. Nevertheless, its clinical relevance, functional roles, and mechanism have not been studied in PDAC. In the current study, real-time quantitative polymerase chain reaction (RT-qPCR) was used to detect the expression of circ_0007534 in 60-paired PDAC tissue samples and different cell lines. Loss-of-function and gain-of-function assays were performed to detect cell proliferation, apoptosis, and metastatic properties affected by circ_0007534. An animal study was also carried out. The luciferase reporter assay was performed to uncover the underlying mechanism of circ_0007534. As a result, circ_0007534 was overexpressed not only in PDAC tissues but also in a panel of PDAC cell lines, and this overexpression is closely associated with advanced tumor stage and positive lymph node invasion. In addition, circ_0007534 may be regarded as an independent prognostic factor for patients with PDAC. For the part of functional assays, circ_0007534 significantly increased cell proliferation, migratory, and invasive potential of PDAC cells. Circ_0007534 could inhibit cell apoptosis partly via a Bcl-2/caspase-3 pathway.The xenograft study further confirmed the cell growth promoting the role of circ_0007534. Mechanistically, miR-625 and miR-892b were sponged by circ_0007534. The oncogenic functions of circ_0007534 is partly dependent on its regulation of miR-625 and miR-892b. In conclusion, our study illuminates a novel circRNA that confers an oncogenic function in PDAC. K E Y W O R D S circ_0007534, circular RNA, miR-625, miR-892b, pancreatic ductal adenocarcinoma J Cell Biochem. 2019;120:3780-3789. wileyonlinelibrary.com/journal/jcb 3780 |

No abstract

No abstract

Deep learning models are increasingly used in mobile applications as critical components. Unlike the program bytecode whose vulnerabilities and threats have been widelydiscussed, whether and how the deep learning models deployed in the applications can be compromised are not well-understood since neural networks are usually viewed as a black box. In this paper, we introduce a highly practical backdoor attack achieved with a set of reverse-engineering techniques over compiled deep learning models. The core of the attack is a neural conditional branch constructed with a trigger detector and several operators and injected into the victim model as a malicious payload. The attack is effective as the conditional logic can be flexibly customized by the attacker, and scalable as it does not require any prior knowledge from the original model. We evaluated the attack effectiveness using 5 state-of-the-art deep learning models and real-world samples collected from 30 users. The results demonstrated that the injected backdoor can be triggered with a success rate of 93.5%, while only brought less than 2ms latency overhead and no more than 1.4% accuracy decrease. We further conducted an empirical study on real-world mobile deep learning apps collected from Google Play. We found 54 apps that were vulnerable to our attack, including popular and securitycritical ones. The results call for the awareness of deep learning application developers and auditors to enhance the protection of deployed models.

No abstract