Due to the widespread availability of implicit feedback (e.g., clicks and purchases), some researchers have endeavored to design recommender systems based on implicit feedback. However, unlike explicit feedback, implicit feedback cannot directly reflect user preferences. Therefore, although more challenging, it is also more practical to use implicit feedback for recommender systems. Traditional collaborative filtering methods such as matrix factorization, which regards user preferences as a linear combination of user and item latent vectors, have limited learning capacities and suffer from data sparsity and the cold-start problem. To tackle these problems, some authors have considered the integration of a deep neural network to learn user and item features with traditional collaborative filtering. However, there is as yet no research combining collaborative filtering and contentbased recommendation with deep learning. In this paper, we propose a novel deep hybrid recommender system framework based on auto-encoders (DHA-RS) by integrating user and item side information to construct a hybrid recommender system and enhance performance. DHA-RS combines stacked denoising auto-encoders with neural collaborative filtering, which corresponds to the process of learning user and item features from auxiliary information to predict user preferences. Experiments performed on the real-world dataset reveal that DHA-RS performs better than state-of-the-art methods.

MicroRNAs (miRNAs) may promote the development and progression of human cancers. Therefore, components of the miRNA biogenesis pathway may play critical roles in human cancer. Single nucleotide polymorphisms (SNPs) or mutations in genes involved in the miRNA biogenesis pathway may alter levels of gene expression, affecting disease susceptibility. Results of previous studies on genetic variants in the miRNA biogenesis pathway and cancer risk were inconsistent. Therefore, a meta-analysis is needed to assess the associations of these genetic variants with human cancer risk. We searched for relevant articles from PubMed, Web of Science, CNKI, and CBM through Jun 21, 2016. In total, 21 case-control articles met all of the inclusion criteria for the study. Significant associations were observed between cancer risk and the DGCR8polymorphism rs417309 G >A (OR 1.22, 95% CI [1.04–1.42]), as well as the DICER1 polymorphism rs1057035 TT (OR 1.13, 95% CI [1.05–1.22]). These SNPs exhibit high potential as novel diagnostic markers. Future studies with larger sample sizes and more refined analyses are needed to shed more light on these findings.

AimThis study was designed to investigate the predictive and prognostic values of serum vascular endothelial growth factor (VEGF) level in non-small cell lung cancer (NSCLC) patients treated with platinum-based chemotherapy.MethodsPatients’ peripheral blood samples were collected prior to chemotherapy and after 1 week of the third cycle of combination chemotherapy. Serum VEGF levels were evaluated through Luminex multiplex technique. Between September 2011 and August 2015, a total of 135 consecutive advanced or recurrent histologically verified NSCLC patients were enrolled in the study. Moreover, all the patients received platinum-based combination chemotherapy as a first-line treatment.ResultsNo significant associations were found between pretreatment serum VEGF levels and clinical characteristics, such as sex (P=0.0975), age (P=0.2522), stage (P=0.1407), lymph node metastasis (P=0.6409), tumor location (P=0.3520), differentiated degree (P=0.5608), pathological (histological) type (P=0.4885), and response to treatment (P=0.9859). The VEGF load per platelet (VEGFPLT) levels were not correlated with sex, age, primary tumor site, and pathological type in NSCLC patients (all P>0.05). The median survival time of progression-free survival (PFS) was 6.407 and 5.29 months in the low and high groups, respectively, when using 280 pg/mL VEGF level as the cutoff point (P=0.024).ConclusionIn conclusion, the serum VEGF levels were found to be a poor prognostic biomarker for the efficacy of platinum-based chemotherapy in terms of PFS, but it was not shown to be a suitable predictive marker for clinical response to platinum-based chemotherapy.

The effects of the microRNA (miRNA) processing genes Gemin3 and Gemin4 on cellular signaling pathways could have a major impact on the risk of cancer. Several studies concerning the association between the Gemin3 rs197412, Gemin4 rs7813 and Gemin4 rs2740348 polymorphisms with cancer susceptibility have been published. The present meta-analysis summarized this evidence and evaluated the precision of these relationships. Relevant studies (published prior to December 16th, 2015) without language restriction were identified using the PubMed, Web of Science and China National Knowledge Infrastructure (CNKI) on-line databases. The data were extracted from the eligible studies and were processed using Stata 12.0 software. Seven studies (2,588 cases and 2,549 controls) indicated that the rs7813 polymorphism was significantly associated with increased cancer risk (TT vs TC + CC, OR = 1.18 95% CI [1.05–1.32]). Six studies (1,314 cases and 1,244 controls) indicated that rs2740348 was associated with an increased cancer risk (GG vs. GC + CC, OR = 1.41 95% CI [1.00–1.83]). However the rs197412 polymorphism was not associated with an increased cancer risk (OR = 0.97 95% CI [0.80–1.19]). Our results suggest that the Gemin4 rs7813 T > C and rs2740348 G > C polymorphisms are associated with cancer susceptibility.

Supplemental Digital Content is available in the text