Recently, product images have gained increasing attention in clothing recommendation since the visual appearance of clothing products has a significant impact on consumers' decision. Most existing methods rely on conventional features to represent an image, such as the visual features extracted by convolutional neural networks (CNN features) and the scale-invariant feature transform algorithm (SIFT features), color histograms, and so on. Nevertheless, one important type of features, the aesthetic features, is seldom considered. It plays a vital role in clothing recommendation since a users' decision depends largely on whether the clothing is in line with her aesthetics, however the conventional image features cannot portray this directly. To bridge this gap, we propose to introduce the aesthetic information, which is highly relevant with user preference, into clothing recommender systems. To achieve this, we first present the aesthetic features extracted by a pre-trained neural network, which is a brain-inspired deep structure trained for the aesthetic assessment task. Considering that the aesthetic preference varies significantly from user to user and by time, we then propose a new tensor factorization model to incorporate the aesthetic features in a personalized manner. We conduct extensive experiments on real-world datasets, which demonstrate that our approach can capture the aesthetic preference of users and significantly outperform several state-of-the-art recommendation methods.

Ankylosing spondylitis (AS) is a type of rheumatic disease characterized by chronic inflammation and pathological osteogenesis in the entheses. Previously, we demonstrated that enhanced osteogenic differentiation of MSC from AS patients (AS-MSC) resulted in pathological osteogenesis, and that during the enhanced osteogenic differentiation course, AS-MSC induced TNF-α-mediated local inflammation. However, whether TNF-α in turn affects AS-MSC remains unknown. Herein, we further demonstrate that a high-concentration TNF-α treatment triggers enhanced directional migration of AS-MSC in vitro and in vivo, which enforces AS pathogenesis. Mechanistically, TNF-α leads to increased expression of ELMO1 in AS-MSC, which is mediated by a METTL14 dependent m6A modification in ELMO1 3′UTR. Higher ELMO1 expression of AS-MSC is found in vivo in AS patients, and inhibiting ELMO1 in SKG mice produces therapeutic effects in this spondyloarthritis model. This study may provide insight into not only the pathogenesis but also clinical therapy for AS.

Hepatitis B virus (HBV), a small enveloped DNA virus, chronically infects more than 350 million people worldwide and causes liver diseases from hepatitis to cirrhosis and liver cancer. Here, we report that hepatocyte nuclear factor 6 (HNF6), a liver-enriched transcription factor, can inhibit HBV gene expression and DNA replication. Overexpression of HNF6 inhibited, while knockdown of HNF6 expression enhanced, HBV gene expression and replication in hepatoma cells. Mechanistically, the SP2 promoter was inhibited by HNF6, which partly accounts for the inhibition on S mRNA. Detailed analysis showed that a cis element on the HBV genome (nucleotides [nt] 3009 to 3019) was responsible for the inhibition of the SP2 promoter by HNF6. Moreover, further analysis showed that HNF6 reduced viral pregenomic RNA (pgRNA) posttranscriptionally via accelerating the degradation of HBV pgRNA independent of La protein. Furthermore, by using truncated mutation experiments, we demonstrated that the N-terminal region of HNF6 was responsible for its inhibitory effects. Importantly, introduction of an HNF6 expression construct with the HBV genome into the mouse liver using hydrodynamic injection resulted in a significant reduction in viral gene expression and DNA replication. Overall, our data demonstrated that HNF6 is a novel host factor that can restrict HBV replication via both transcriptional and posttranscriptional mechanisms. IMPORTANCEHBV is a major human pathogen whose replication is regulated by host factors. Liver-enriched transcription factors are critical for many liver functions, including metabolism, development, and cell proliferation, and some of them have been shown to regulate HBV gene expression or replication in different manners. In this study, we showed that HNF6 could inhibit the gene expression and DNA replication of HBV via both transcriptional and posttranscriptional mechanisms. As HNF6 is differentially expressed in men and women, the current results may suggest a role of HNF6 in the gender dimorphism of HBV infection.

Implicit feedback data is extensively explored in recommendation as it is easy to collect and generally applicable. However, predicting users' preference on implicit feedback data is a challenging task since we can only observe positive (voted) samples and unvoted samples. It is difficult to distinguish between the negative samples and unlabeled positive samples from the unvoted ones. Existing works, such as Bayesian Personalized Ranking (BPR), sample unvoted items as negative samples uniformly, therefore suffer from a critical noisy-label issue. To address this gap, we design an adaptive sampler based on noisy-label robust learning for implicit feedback data. To formulate the issue, we first introduce Bayesian Point-wise Optimization (BPO) to learn a model, e.g., Matrix Factorization (MF), by maximum likelihood estimation. We predict users' preferences with the model and learn it by maximizing likelihood of observed data labels, i.e., a user prefers her positive samples and has no interests in her unvoted samples. However, in reality, a user may have interests in some of her unvoted samples, which are indeed positive samples mislabeled as negative ones. We then consider the risk of these noisy labels, and propose a Noisy-label Robust BPO (NBPO). NBPO also maximizes the observation likelihood while connects users' preference and observed labels by the likelihood of label flipping based on the Bayes' theorem. In NBPO, a user prefers her true positive samples and shows no interests in her true negative samples, hence the optimization quality is dramatically improved. Extensive experiments on two public real-world datasets show the significant improvement of our proposed optimization methods.

BackgroundThe relevance of recurrent molecular abnormalities in cytogenetically normal (CN) acute myeloid leukemia (AML) was recently acknowledged by the inclusion of molecular markers such as NPM1, FLT3, and CEBPA as a complement to cytogenetic information within both the World Health Organization and the European Leukemia Net classifications. Mitochondrial metabolism is different in cancer and normal cells. A novel cytosolic type 2-hydroxybutyrate dehydrogenase, BDH2, originally named DHRS6, plays a physiological role in the cytosolic utilization of ketone bodies, which can subsequently enter mitochondria and the tricarboxylic acid cycle. Moreover, BDH2 catalyzes the production of 2, 3-DHBA during enterobactin biosynthesis and participates in 24p3 (LCN2)-mediated iron transport and apoptosis.ResultsWe observed that BDH2 expression is an independent poor prognostic factor for CN-AML, with an anti-apoptotic role. Patients with high BDH2 expression have relatively shorter overall survival (P = 0.007) and a low complete response rate (P = 0.032). BDH2-knockdown (BDH2-KD) in THP1 and HL60 cells increased the apoptosis rate under reactive oxygen species stimulation. Decrease inducible survivin, a member of the inhibitors of apoptosis family, but not members of the Bcl-2 family, induced apoptosis via a caspase-3-independent pathway upon BDH2-KD.ConclusionsBDH2 is a novel independent poor prognostic marker for CN-AML, with the role of anti-apoptosis, through surviving.