The application of deep learning approaches to finance has received a great deal of attention from both investors and researchers. This study presents a novel deep learning framework where wavelet transforms (WT), stacked autoencoders (SAEs) and long-short term memory (LSTM) are combined for stock price forecasting. The SAEs for hierarchically extracted deep features is introduced into stock price forecasting for the first time. The deep learning framework comprises three stages. First, the stock price time series is decomposed by WT to eliminate noise. Second, SAEs is applied to generate deep high-level features for predicting the stock price. Third, high-level denoising features are fed into LSTM to forecast the next day’s closing price. Six market indices and their corresponding index futures are chosen to examine the performance of the proposed model. Results show that the proposed model outperforms other similar models in both predictive accuracy and profitability performance.

Breast cancer (BC) is the most common female malignancy within the spectrum of human cancer. One promising way to reduce the mortality and morbidity of BC is to explore novel diagnostic markers for early diagnosis and prognostication. The neutrophil lymphocyte ratio (NLR) is a good reflection of inflammation, which plays an important role in tumor progression and metastasis. However, the association between NLR and BC prognosis remains unclear. The aim of this meta-analysis is to explore the prognostic value of NLR in BC. Among the screened references in the database, 12 eligible studies were identified in this study. Patients with a higher NLR had a shorter disease-free survival (hazard ratio =1.46, 95% confidence interval: 1.12–1.90, P=0.044) and overall survival (hazard ratio =2.03, 95% confidence interval: 1.41–2.93, P<0.001). In the subgroup analysis of NLR and disease-free survival, the studies from Eastern countries had a positive result with perfect homogeneity (I2=0); however, this homogeneity has not been achieved in studies from Western countries. In the subgroup analysis of the NLR and overall survival, the results of the univariate and multivariate analyses were completely different, with different heterogeneity. In the luminal A and luminal B subtypes, we found that there was no association between the NLR and overall survival in the BC patients. Positive results were obtained in the analyses of the human epidermal growth factor receptor 2 (HER2)-positive and triple-negative BC subtypes. In conclusion, this meta-analysis suggests that NLR is a good prognostic marker for BC, and patients with a higher NLR have poorer prognoses. Future studies should perform more detailed investigations to decrease heterogeneity and determine the appropriate cut-off values for different races.

Graphical Abstract Highlights d OTUB2 enhances metastasis through Hippo-independent activation of YAP/TAZ signaling d Poly-SUMOylation is required for OTUB2 to interact with and deubiquitinate YAP/TAZ d YAP/TAZ contain a SUMO-interacting motif (SIM) required for binding SUMOylated OTUB2 d EGF treatment or KRAS mutation stabilizes YAP/TAZ by stimulating OTUB2 SUMOylation In Brief Zhang et al. identified OTUB2 as a cancer stemness and metastasis-promoting factor that deubiquitinates and activates YAP/TAZ independent of the Hippo signaling. OTUB2 was induced by EGF/KRAS to be poly-SUMOylated on lysine 233, and this SUMOylation enables it to bind YAP/TAZ through a yetunknown SUMO-interacting motif (SIM). SUMMARY The transcriptional regulators YAP and TAZ play important roles in development, physiology, and tumorigenesis and are negatively controlled by the Hippo pathway. It is yet unknown why the YAP/ TAZ proteins are frequently activated in human malignancies in which the Hippo pathway is still active.Here, by a gain-of-function cancer metastasis screen, we discovered OTUB2 as a cancer stemness and metastasis-promoting factor that deubiquitinates and activates YAP/TAZ. We found OTUB2 to be poly-SUMOylated on lysine 233, and this SUMOylation enables it to bind YAP/TAZ. We also identified a yetunknown SUMO-interacting motif (SIM) in YAP and TAZ required for their association with SUMOylated OTUB2. Importantly, EGF and oncogenic KRAS induce OTUB2 poly-SUMOylation and thereby activate YAP/TAZ. Our results establish OTUB2 as an essential modulator of YAP/TAZ and also reveal a novel mechanism via which YAP/TAZ activity is induced by oncogenic KRAS.

Cancer-associated inflammation is a key determinant of disease progression and survival in most cancers. The aim of our study was to assess the predictive value of preoperative inflammatory markers, such as the neutrophil–lymphocyte ratio (NLR), platelet–lymphocyte ratio, red cell distribution width (RDW), and mean platelet volume, for survival in breast cancer patients. In total, 608 breast cancer patients operated on between January 2009 and December 2011 were included in this observational study. The association between preoperative inflammatory markers and survival outcomes was analyzed. Patients with high NLR (>2.57) or high RDW (>13.45%) showed a significantly lower overall survival rate than those with lower NLR (≤2.57) or lower RDW (≤13.45%). NLR and RDW, along with node stage and molecular subtypes, were independent prognostic factors. There was a significant survival difference according to NLR in the luminal A and triple-negative subtypes (93.3% versus 99.3%, P=0.001; 68.8% versus 95.1%, P=0.000, respectively). The triple-negative subtype was the only subtype in which higher RDW patients showed significantly poor prognosis (81.3% versus 95.5%, P=0.025). Pre-operation NLR and RDW is a convenient, easily measured prognostic indicator for patients with breast cancer, especially in patients with the triple-negative subtype.

BackgroundIntratumoral and circulating regulatory T cells (Tregs) have been shown to be critical in the pathogenesis of hepatocellular carcinoma (HCC). However there is limited knowledge on the alterations of regulatory B cells (Bregs). We here investigated perioperative dynamic alterations of peripheral circulating Tregs and Bregs in HCC patients to reveal the relationship between regulatory lymphocytes and its clinical implications.Methods36 patients with HCC, 6 with chronic hepatitis B infection and 10 healthy donors were enrolled for this study. Frequencies of peripheral Tregs and Bregs were measured by flow cytometry with antibodies against CD4, CD25, CD127, CD19 and IL-10 before, and after radical surgery. Then, clinical informatics of HCC patients was achieved through Digital Evaluation Score System (DESS) for the assessment of disease severity. Finally, we analysed correlations between digitalized clinical features and kinetics of circulating regulatory lymphocytes.ResultsLevel of circulating CD4+CD25+CD127- Tregs in HCC patients was significantly lower than that in healthy donors and patients with chronic hepatitis B infection before surgery, but was increased after surgery. Preoperative level of CD19+ IL-10+ Bregs in HCC patients was also significantly lower than the other groups. However it dramatically was elevated right after surgery and remained elevated compared to controls (about 7 days after surgery, P = 0.04). Frequency of circulating Tregs was correlated with circulating leukocytes, ferritin, and clinical features suggesting tumor aggressiveness including portal vein thrombosis, hepatic vein involvement and advanced clinical stages. Frequency of circulating Bregs was associated with Hepatitis B e Antigen (HBeAg) and Hepatitis B virus (HBV) DNA copy number. In addition, DESS was significantly and positively correlated with other staging systems.ConclusionFrequencies of peripheral Tregs and Bregs in HCC patients increased after surgery. These results suggest that a postoperative combination of therapies against Tregs and Bregs may be beneficial for better outcome of HCC patients after resection.