Fibroblasts can be reprogrammed to induced pluripotent stem cells (iPSCs) by application of transcription factors octamer-binding protein 4 (Oct4), SRY-box containing gene 2 (Sox2), Kruppel-like factor 4 (Klf4), and c-Myelocytomatosis oncogene (c-Myc) (OSKM), but the underlying mechanisms remain unclear. Here, we report that exog-

Background:The role of the lysine acetyltransferase GCN5 in cancer development remains largely unknown. Results: GCN5 expression correlates with lung cancer tumor size, directly enhances the expression of E2F1, cyclin E1, and cyclin D1, and potentiates lung cancer growth. Conclusion: GCN5 potentiates lung cancer growth in an E2F1-dependent manner. Significance: GCN5 is critical for lung cancer growth and represents a potential target for the treatment of lung cancer.

No abstract

Leukemia inhibitory factor/Stat3 signaling is critical for maintaining the self-renewal and differentiation potential of mouse embryonic stem cells (mESCs). However, the upstream effectors of this pathway have not been clearly defined. Here, we show that periodic tryptophan protein 1 (Pwp1), a WD-40 repeat-containing protein associated with histone H4 modification, is required for the exit of mESCs from the pluripotent state into all lineages. Knockdown (KD) of Pwp1 does not affect mESC proliferation, self-renewal, or apoptosis. However, KD of Pwp1 impairs the differentiation potential of mESCs both in vitro and in vivo. PWP1 chromatin immunoprecipitation-seq results revealed that the PWP1-occupied regions were marked with significant levels of H4K20me3. Moreover, Pwp1 binds to sites in the upstream region of Stat3. KD of Pwp1 decreases the level of H4K20me3 in the upstream region of Stat3 gene and upregulates the expression of Stat3. Furthermore, Pwp1 KD mESCs recover their differentiation potential through suppressing the expression of Stat3 or inhibiting the tyrosine phosphorylation of STAT3. Together, our results suggest that Pwp1 plays important roles in the differentiation potential of mESCs. STEM CELLS 2015;33:661-673

The maturation of induced pluripotent stem cells (iPS) is one of the limiting steps of somatic cell reprogramming, but the underlying mechanism is largely unknown. Here, we reported that knockdown of histone deacetylase 2 (HDAC2) specifically promoted the maturation of iPS cells. Further studies showed that HDAC2 knockdown significantly increased histone acetylation, facilitated TET1 binding and DNA demethylation at the promoters of iPS cell maturation-related genes during the transition of pre-iPS cells to a fully reprogrammed state. We also found that HDAC2 competed with TET1 in the binding of the RbAp46 protein at the promoters of maturation genes and knockdown of TET1 markedly prevented the activation of these genes. Collectively, our data not only demonstrated a novel intrinsic mechanism that the HDAC2-TET1 switch critically regulates iPS cell maturation, but also revealed an underlying mechanism of the interplay between histone acetylation and DNA demethylation in gene regulation.

No abstract

With the rapid progress of urbanization, predicting citywide crowd flows has become increasingly significant in many fields, such as traffic management and public security. However, influenced by the complex spatiotemporal relations in raw data and other factors, such as events and weather, obtaining a precise prediction is challenging. Some previous works attempted to address this problem using various ways, such as autoregressive integrated moving average, vector auto-regression and some deep learning models. However, seldom can these methods comprehensively capture the spatiotemporal correlations. In this paper, we propose a novel spatio-temporal prediction model that is based on densely connected convolutional networks and attention long short-term memory (ST-DCCNAL), to simultaneously predict the inflow and outflow of the crowds in regions divided within a specific city. The ST-DCCNAL model consists of three parts: spatial part, external factors part and temporal part. In the spatial part, we employ densely connected convolutional networks to extract spatial characteristics at different levels. The external factors part utilizes a fully connected network to extract features from auxiliary information. In the last part, an attention-based long short-term memory module is leveraged to capture the temporal pattern. To demonstrate the practicality and effectiveness of the proposed model, we evaluate it using two separate real-world datasets of taxis in Beijing and bikes in New York. The experimental results confirm that the performance of our model is better than that of other baseline methods.INDEX TERMS Data mining, spatiotemporal modeling, crowd flow prediction, densely connected convolutional network, long short-term memory, attention mechanism.

The extrapolation strategy raised by Nesterov, which can accelerate the convergence rate of gradient descent methods by orders of magnitude when dealing with smooth convex objective, has led to tremendous success in training machine learning tasks. In this article, the convergence of individual iterates of projected subgradient (PSG) methods for nonsmooth convex optimization problems is theoretically studied based on Nesterov's extrapolation, which we name individual convergence. We prove that Nesterov's extrapolation has the strength to make the individual convergence of PSG optimal for nonsmooth problems. In light of this consideration, a direct modification of the subgradient evaluation suffices to achieve optimal individual convergence for strongly convex problems, which can be regarded as making an interesting step toward the open question about stochastic gradient descent (SGD) posed by Shamir. Furthermore, we give an extension of the derived algorithms to solve regularized learning tasks with nonsmooth losses in stochastic settings. Compared with other state-of-theart nonsmooth methods, the derived algorithms can serve as an alternative to the basic SGD especially in coping with machine learning problems, where an individual output is needed to guarantee the regularization structure while keeping an optimal rate of convergence. Typically, our method is applicable as an efficient tool for solving large-scale l 1 -regularized hinge-loss learning problems. Several comparison experiments demonstrate that our individual output not only achieves an optimal convergence rate but also guarantees better sparsity than the averaged solution.