Search

Purpose Currently, the routine screening program has insufficient capacity for the early diagnosis of lung cancer. Therefore, a type of chitosan-molecular beacon (CS-MB) probe was developed to recognize the miR-155-5p and image the lung cancer cells for the early diagnosis. Methods Based on the molecular beacon (MB) technology and nanotechnology, the CS-MB probe was synthesized self-assembly. There are four types of cells—three kinds of animal models and one type of histopathological sections of human lung cancer were utilized as models, including A549, SPC-A1, H446 lung cancer cells, tumor-initiating cells (TICs), subcutaneous and lung xenografts mice, and lox-stop-lox(LSL) K-ras G12D transgenic mice. The transgenic mice dynamically displayed the process from normal lung tissues to atypical hyperplasia, adenoma, carcinoma in situ, and adenocarcinoma. The different miR-155-5p expression levels in these cells and models were measured by quantitative real-time polymerase chain reaction (qRT-PCR). The CS-MB probe was used to recognize the miR-155-5p and image the lung cancer cells by confocal microscopy in vitro and by living imaging system in vivo. Results The CS-MB probe could be used to recognize the miR-155-5p and image the lung cancer cells significantly in these cells and models. The fluorescence intensity trends detected by the CS-MB probe were similar to the expression levels trends of miR-155 tested by qRT-PCR. Moreover, the fluorescence intensity showed an increasing trend with the tumor progression in the transgenic mice model, and the occurrence and development of lung cancer were dynamically monitored by the differen fluorescence intensity. In addition, the miR-155-5p in human lung cancer tissues could be detected by the miR-155-5p MB. Conclusion Both in vivo and in vitro experiments demonstrated that the CS-MB probe could be utilized to recognize the miR-155-5p and image the lung cancer cells. It provided a novel experimental and theoretical basis for the early diagnosis of the disease. Also, the histopathological sections of human lung cancer research laid the foundation for subsequent preclinical studies. In addition, different MBs could be designed to detect other miRNAs for the early diagnosis of other tumors.

Objective To explore the effect of miR-182 in non small cell lung cancer (NSCLC) and its mechanism. Methods The expression level and difference of miR-182 were detected by qRT-PCR in NSCLC cell line and matched tissues of 30 patients suffering from NSCLC. Cell proliferation, invasion and metastasis of A549 and H1299 via miR-182 was detected by miR-182 mimics. The predicted target gene (RECK) of miR-182 was identified via luciferase activation assay. The mRNA or protein expression of RECK after miR-182 mimic or inhibitor transfection was detected by qRT-PCR or Western blotting, respectively. Results Expression of miR-182 was significantly increased in NSCLC tissues and cell lines compared with that in the adjacent tissues (P

Mining closed frequent item set(CFI) plays a fundamental role in many real-world data mining applications. However, memory requirement and computational cost have become the bottleneck of CFI mining algorithms, particularly when confronting with large scale datasets, which herewith makes mining closed frequent item set from large scale datasets a significant and challenging issue. To address the above issue, a parallelized AFOPT-close algorithm is proposed and implemented in this paper based on the cloud computing framework MapReduce, which is widely used to cope with large scale data. Furthermore, an efficient parallelized method for checking if a frequent item set is globally closed is also proposed on the MapReduce platform to further improve the mining performance. Experimental results are then provided and analyzed to verify the efficiency and effectiveness of the proposed methods for mining closed frequent item set.

Mining closed frequent itemset (CFI) plays an essential role in many real-world data mining applications. With the emergence of abundant large-scale data sets, it now turns to be a significant and challenging issue to mine CFI concurrently. This paper proposes a parallel balanced mining algorithm for CFI based on the MapReduce platform. The proposed algorithm adopts Greedy strategy to group items aiming to balance the computation burdens among all parallel tasks, which is consisted of three main steps: (1) Parallel Counting, (2) Global Construction of Frequent List (F_list) and Group Map (G_map), (3) Parallel Mining for Closed Frequent Itemset. Experimental results validate the method and show its effectiveness as satisfied speedup and scalability are both achieved in large-scale CFI mining tasks.