Your browser doesn't support javascript.
Mostrar: 20 | 50 | 100
Resultados 1 - 20 de 374
Filtrar
1.
Sci Adv ; 6(12): eaaw5790, 2020 03.
Artigo em Inglês | MEDLINE | ID: mdl-32206705

RESUMO

Atmospheric nitrogen (N) deposition affects the greenhouse gas (GHG) balance of ecosystems through the net atmospheric CO2 exchange and the emission of non-CO2 GHGs (CH4 and N2O). We quantified the effects of N deposition on biomass increment, soil organic carbon (SOC), and N2O and CH4 fluxes and, ultimately, the net GHG budget at ecosystem level of a Moso bamboo forest in China. Nitrogen addition significantly increased woody biomass increment and SOC decomposition, increased N2O emission, and reduced soil CH4 uptake. Despite higher N2O and CH4 fluxes, the ecosystem remained a net GHG sink of 26.8 to 29.4 megagrams of CO2 equivalent hectare-1 year-1 after 4 years of N addition against 22.7 hectare-1 year-1 without N addition. The total net carbon benefits induced by atmospheric N deposition at current rates of 30 kilograms of N hectare-1 year-1 over Moso bamboo forests across China were estimated to be of 23.8 teragrams of CO2 equivalent year-1.

2.
Artigo em Inglês | MEDLINE | ID: mdl-32217486

RESUMO

Deep neural networks (DNNs) thrive in recent years, wherein batch normalization (BN) plays an indispensable role. However, it has been observed that BN is costly due to the huge reduction and elementwise operations that are hard to be executed in parallel, which heavily reduces the training speed. To address this issue, in this article, we propose a methodology to alleviate the BN's cost by using only a few sampled or generated data for mean and variance estimation at each iteration. The key challenge to reach this goal is how to achieve a satisfactory balance between normalization effectiveness and execution efficiency. We identify that the effectiveness expects less data correlation in sampling while the efficiency expects more regular execution patterns. To this end, we design two categories of approach: sampling or creating a few uncorrelated data for statistics' estimation with certain strategy constraints. The former includes ``batch sampling (BS)'' that randomly selects a few samples from each batch and ``feature sampling (FS)'' that randomly selects a small patch from each feature map of all samples, and the latter is ``virtual data set normalization (VDN)'' that generates a few synthetic random samples to directly create uncorrelated data for statistics' estimation. Accordingly, multiway strategies are designed to reduce the data correlation for accurate estimation and optimize the execution pattern for running acceleration in the meantime. The proposed methods are comprehensively evaluated on various DNN models, where the loss of model accuracy and the convergence rate are negligible. Without the support of any specialized libraries, 1.98x BN layer acceleration and 23.2% overall training speedup can be practically achieved on modern GPUs. Furthermore, our methods demonstrate powerful performance when solving the well-known ``micro-BN'' problem in the case of a tiny batch size. This article provides a promising solution for the efficient training of high-performance DNNs.

3.
J Chem Inf Model ; 2020 Mar 13.
Artigo em Inglês | MEDLINE | ID: mdl-32118415

RESUMO

Drug research and development is a time-consuming and high-cost task, pressing an urgent demand to identify novel indications of approved drugs, referred to as drug repositioning, which provides an economical and efficient way for drug discovery. With increasing volumes of large-scale chemical, genomic, and pharmacological data sets generated by the high-throughput technique, it is crucial to develop systematic and rational computational approaches to identify new indications of approved drugs. In this paper, we introduce HNet-DNN, which utilizes a deep neural network (DNN), to predict new drug-disease associations based on the features extracted from the drug-disease heterogeneous network. Instead of the straightforward concatenation of chemical and phenotypic features as the input of DNN, we used these raw features of drugs and diseases to construct a drug-drug similarity network and a disease-disease similarity network, and then built a drug-disease heterogeneous network by integrating known drug-disease associations. Subsequently, we extracted topological features for drug-disease associations from the heterogeneous network and used them to train a DNN model. Our intensive performance evaluations demonstrated that HNet-DNN effectively exploits the features of the heterogeneous network to boost the predictive performance of drug-disease associations. Compared with a couple of typical classifiers and competitive approaches, our method not only achieved state-of-the-art performance but also effectively alleviated the overfitting problem. Moreover, we ran HNet-DNN to predict new drug-disease associations and carried out case studies to verify the effectiveness of our method.

4.
Phys Chem Chem Phys ; 22(10): 5548-5560, 2020 Mar 11.
Artigo em Inglês | MEDLINE | ID: mdl-32119016

RESUMO

The entry of human immunodeficiency virus type I (HIV-1) into host cells is initiated by binding to the cell-surface receptor CD4, which induces a conformational transition of the envelope (Env) glycoprotein gp120 from the closed, unliganded state to the open, CD4-bound state. Despite many available structures in these two states, detailed aspects on the dynamics and thermodynamics of gp120 remain elusive. Here, we performed microsecond-scale (µs-scale) multiple-replica molecular dynamics (MD) simulations to explore the differences in the conformational dynamics, protein motions, and thermodynamics between the unliganded and CD4-bound/complexed forms of gp120. Comparative analyses of MD trajectories reveal that CD4 binding promotes the structural deviations/changes and conformational flexibility, loosens the structural packing, and complicates the molecular motions of gp120. Comparison of the constructed free energy landscapes (FELs) reveals that the CD4-complexed gp120 has more conformational substates, larger conformational entropy, and lower thermostability than the unliganded form. Therefore, the unliganded conformation represents a structurally and energetically stable "ground state" for the full-length gp120. The observed great increase in the mobility of V1/V2 and V3 along with their more versatile movement directions in the CD4-bound gp120 compared to the unliganded form suggests that their orientations with respect to each other and to the structural core determine the differences in the conformational dynamics and thermodynamics between the two gp120 forms. The results presented here provide a basis by which to better understand the functional and immunological properties of gp120 and, furthermore, to deploy appropriate strategies for the development of anti-HIV-1 drugs or vaccines.


Assuntos
Antígenos CD4/metabolismo , Proteína gp120 do Envelope de HIV/metabolismo , Simulação de Dinâmica Molecular , Termodinâmica , Ligantes , Ligação Proteica , Conformação Proteica
5.
Neural Netw ; 125: 70-82, 2020 Jan 15.
Artigo em Inglês | MEDLINE | ID: mdl-32070857

RESUMO

Deep neural network (DNN) quantization converting floating-point (FP) data in the network to integers (INT) is an effective way to shrink the model size for memory saving and simplify the operations for compute acceleration. Recently, researches on DNN quantization develop from inference to training, laying a foundation for the online training on accelerators. However, existing schemes leaving batch normalization (BN) untouched during training are mostly incomplete quantization that still adopts high precision FP in some parts of the data paths. Currently, there is no solution that can use only low bit-width INT data during the whole training process of large-scale DNNs with acceptable accuracy. In this work, through decomposing all the computation steps in DNNs and fusing three special quantization functions to satisfy the different precision requirements, we propose a unified complete quantization framework termed as "WAGEUBN" to quantize DNNs involving all data paths including W (Weights), A (Activation), G (Gradient), E (Error), U (Update), and BN. Moreover, the Momentum optimizer is also quantized to realize a completely quantized framework. Experiments on ResNet18/34/50 models demonstrate that WAGEUBN can achieve competitive accuracy on the ImageNet dataset. For the first time, the study of quantization in large-scale DNNs is advanced to the full 8-bit INT level. In this way, all the operations in the training and inference can be bit-wise operations, pushing towards faster processing speed, decreased memory cost, and higher energy efficiency. Our throughout quantization framework has great potential for future efficient portable devices with online learning ability.

6.
Oncologist ; 2020 Feb 21.
Artigo em Inglês | MEDLINE | ID: mdl-32083766

RESUMO

BACKGROUND: The role of postoperative radiotherapy in pathological T2-3N0M0 esophageal squamous cell carcinoma is unknown. We aimed to evaluate the efficacy and safety of postoperative radiotherapy in patients with pathological T2-3N0M0 thoracic esophageal squamous cell carcinoma. MATERIALS AND METHODS: Patients aged 18-72 years with pathological stage T2-3N0M0 esophageal squamous cell carcinoma after radical surgery and without neoadjuvant therapy were eligible. Patients were randomly assigned to surgery alone or to receive postoperative radiotherapy of 50.4 Gy in supraclavicular field and 56 Gy in mediastinal field in 28 fractions over 6 weeks. The primary endpoint was disease-free survival. The secondary endpoints were local-regional recurrence rate, overall survival, and radiation-related toxicities. RESULTS: From October 2012 to February 2018, 167 patients were enrolled in this study. We analyzed 157 patients whose follow-up time was more than 1 year or who had died. The median follow-up time was 45.6 months. The 3-year disease-free survival rates were 75.1% (95% confidence interval [CI] 65.9-85.5) in the postoperative radiotherapy group and 58.7% (95% CI 48.2-71.5) in the surgery group (hazard ratio 0.53, 95% CI 0.30-0.94, p = .030). Local-regional recurrence rate decreased significantly in the radiotherapy group (10.0% vs. 32.5% in the surgery group, p = .001). The overall survival and distant metastasis rates were not significantly different between two groups. Grade 3 toxicity rate related to radiotherapy was 12.5%. CONCLUSION: Postoperative radiotherapy significantly increased disease-free survival and decreased local regional recurrence rate in patients with pathological T2-3N0M0 thoracic esophageal squamous cell carcinoma with acceptable toxicities in this interim analysis. Further enrollment and follow-up are warranted to validate these findings in this ongoing trial. IMPLICATIONS FOR PRACTICE: The value of adjuvant radiotherapy for patients with node-negative esophageal cancer is not clear. The interim results of this phase III study indicated that postoperative radiotherapy significantly improved disease-free survival and decreased local-regional recurrence rate in patients with pathological T2-3N0M0 thoracic esophageal squamous cell carcinoma compared with surgery alone with acceptable toxicities. The distant metastasis rates and overall survival rates were not different between the two groups. Adjuvant radiotherapy should be considered for pathologic T2-3N0M0 thoracic esophageal squamous cell carcinoma. Prospective trials to identify high-risk subgroups are needed.

7.
BMC Cancer ; 20(1): 144, 2020 Feb 22.
Artigo em Inglês | MEDLINE | ID: mdl-32087687

RESUMO

BACKGROUND: Information on the optimal salvage regimen for recurrent esophageal cancer is scarce. We aimed to assess the patterns of locoregional failure, and evaluate the therapeutic efficacy of salvage therapy along with the prognostic factors in recurrent thoracic esophageal squamous cell carcinoma (TESCC) after radical esophagectomy. METHODS: A total of 193 TESCC patients who were diagnosed with recurrence after radical surgery and received salvage treatment at our hospital were retrospectively reviewed from 2004 to 2014. The patterns of the first failure were assessed. The post-recurrence survival rate was determined using the Kaplan-Meier method and analyzed using the log-rank test. Multivariate prognostic analysis was performed using the Cox proportional hazard model. RESULTS: The median time of failure was 7.0 months. Among the 193 patients, 163 exhibited isolated locoregional lymph node (LN) recurrence and 30 experienced locoregional LN relapse with hematogenous metastasis. Among the 193 patients, LN recurrence was noted at 302 sites; the most common sites included the supraclavicular (25.8%; 78/302) and mediastinal LNs (44.4%; 134/302), particularly stations 1 to 6 for the mediastinal LNs (36.4%; 110/302). The median overall survival (OS) was 13.1 months after recurrence. In those treated with salvage chemoradiotherapy, with radiotherapy, and without radiotherapy, the 1-year OS rates were 68.5, 55.0, and 28.6%; the 3-year OS rates were 35.4, 23.8, and 2.9%; and the 5-year OS rates were 31.8, 17.2, 2.9%, respectively (P < 0.001). Furthermore, patient survival in those who received salvage chemoradiotherapy was significantly better than those treated with salvage radiotherapy alone (P = 0.044). Multivariate analysis showed that the pathological TNM stage and salvage treatment regimen were independent prognostic factors. CONCLUSIONS: Supraclavicular and mediastinal LN failure were the most common types of recurrence after R0 surgery in TESCC patients. Salvage chemoradiotherapy or radiotherapy could significantly improve survival in esophageal cancer with locoregional LN recurrence.

8.
Biochim Biophys Acta Biomembr ; 1862(6): 183217, 2020 Feb 13.
Artigo em Inglês | MEDLINE | ID: mdl-32061646

RESUMO

As the only exposed viral protein at the membrane surface of HIV, envelope glycoprotein gp120 is responsible for recognizing host cells and mediating virus-cell membrane fusion. Available structures of gp120 indicate that it exhibits two distinct conformational states, called closed and open states. Although experimental data demonstrates that CD4 binding stabilizes open state of gp120, detailed structural dynamics and kinetics of gp120 during this process remain elusive. Here, two open-state gp120 simulation systems, one without any ligands (ligand-free) and the other complexed with CD4 (CD4-bound), were subjected to microsecond-scale molecular dynamics simulations following the conformational transitions and allosteric pathways of gp120 evaluated by using the Markov state model and a network-based method, respectively. Our results provide an atomic-resolution description of gp120 conformational transitions, suggesting that gp120 is intrinsically dynamic from the open state to closed state, whereas CD4 binding blocks these transitions. Consistent with experimental structures, five metastable conformations with different orientations of the V1/V2 region and V3 loop have been extracted. The binding of CD4 significantly enhances allosteric communications from the CD4-binding site to V3 loop and ß20-21 hairpin, resulting in high-affinity interactions with coreceptors and activation of the conformational transitions switcher, respectively. This study will facilitate the structural understanding of the CD4-binding effects on conformational transitions and allosteric pathways of gp120.

9.
Sci Rep ; 10(1): 1026, 2020 Jan 23.
Artigo em Inglês | MEDLINE | ID: mdl-31974403

RESUMO

Cryptosporidium spp. and Enterocytozoon bieneusi are two well-known protist pathogens which can result in diarrhea in humans and animals. To examine the occurrence and genetic characteristics of Cryptosporidium spp. and E. bieneusi in pet red squirrels (Sciurus vulgaris), 314 fecal specimens were collected from red squirrels from four pet shops and owners in Sichuan province, China. Cryptosporidium spp. and E. bieneusi were examined by nested PCR targeting the partial small subunit rRNA (SSU rRNA) gene and the ribosomal internal transcribed spacer (ITS) gene respectively. The infection rates were 8.6% (27/314) for Cryptosporidium spp. and 19.4% (61/314) for E. bieneusi. Five Cryptosporidium species/genotypes were identified by DNA sequence analysis: Cryptosporidium rat genotype II (n = 8), Cryptosporidium ferret genotype (n = 8), Cryptosporidium chipmunk genotype III (n = 5), Cryptosporidium rat genotype I (n = 4), and Cryptosporidium parvum (n = 2). Additionally, a total of five E. bieneusi genotypes were revealed, including three known genotypes (D, SCC-2, and SCC-3) and two novel genotypes (RS01 and RS02). Phylogenetic analysis revealed that genotype D fell into group 1, whereas the remaining genotypes clustered into group 10. To our knowledge, this is the first study to report Cryptosporidium spp. and E. bieneusi in pet red squirrels in China. Moreover, C. parvum and genotype D of E. bieneusi, previously identified in humans, were also found in red squirrels, suggesting that red squirrels may give rise to cryptosporidiosis and microsporidiosis in humans through zoonotic transmissions. These results provide preliminary reference data for monitoring Cryptosporidium spp. and E. bieneusi infections in pet red squirrels and humans.

10.
Curr Pharm Des ; 2020 Jan 16.
Artigo em Inglês | MEDLINE | ID: mdl-31951162

RESUMO

Computational drug repositioning is an efficient approach towards discovering new indications for existing drugs. In recent years, with the accumulation of online health-related information and the extensive use of biomedical databases, computational drug repositioning approaches have achieved significant progress in drug discovery. In this review, we summarize recent advancements in drug repositioning. Firstly, we detailed demonstrate available data source information which is conducive to identifying novel indications. Furthermore, we provide a summary of commonly used computing approaches. For each method, we briefly describe techniques, case studies, and evaluation criteria. Finally, we discuss the limitations of existing computing approaches.

11.
Sci Rep ; 10(1): 1278, 2020 Jan 28.
Artigo em Inglês | MEDLINE | ID: mdl-31992738

RESUMO

The interaction between protein and DNA plays an essential function in various critical natural processes, like DNA replication, transcription, splicing, and repair. Studying the binding affinity of proteins to DNA helps to understand the recognition mechanism of protein-DNA complexes. Since there are still many limitations on the protein-DNA binding affinity data measured by experiments, accurate and reliable calculation methods are necessarily required. So we put forward a computational approach in this paper, called PreDBA, that can forecast protein-DNA binding affinity effectively by using heterogeneous ensemble models. One hundred protein-DNA complexes are manually collected from the related literature as a data set for protein-DNA binding affinity. Then, 52 sequence and structural features are obtained. Based on this, the correlation between these 52 characteristics and protein-DNA binding affinity is calculated. Furthermore, we found that the protein-DNA binding affinity is affected by the DNA molecule structure of the compound. We classify all protein-DNA compounds into five classifications based on the DNA structure related to the proteins that make up the protein-DNA complexes. In each group, a stacked heterogeneous ensemble model is constructed based on the obtained features. In the end, based on the binding affinity data set, we used the leave-one-out cross-validation to evaluate the proposed method comprehensively. In the five categories, the Pearson correlation coefficient values of our recommended method range from 0.735 to 0.926. We have demonstrated the advantages of the proposed method compared to other machine learning methods and currently existing protein-DNA binding affinity prediction approach.

12.
Plant J ; 2020 Jan 10.
Artigo em Inglês | MEDLINE | ID: mdl-31923323

RESUMO

Incorporating male sterility into hybrid seed production reduces its cost and ensures high varietal purity. Despite these advantages, male-sterile lines have not been widely used to produce tomato (Solanum lycopersicum) hybrid seeds. We describe the development of a biotechnology-based breeding platform that utilized genic male sterility to produce hybrid seeds. In this platform, we generated a novel male-sterile tomato line by clustered regularly interspaced short palindromic repeat (CRISPR)/CRISPR-associated protein 9 (Cas9)-mediated mutagenesis of a stamen-specific gene SlSTR1 and devised a transgenic maintainer by transforming male-sterile plants with a fertility-restoration gene linked to a seedling-colour gene. Offspring of crosses between a hemizygous maintainer and the homozygous male-sterile plant segregated into 50% non-transgenic male-sterile plants and 50% male-fertile maintainer plants, which could be easily distinguished by seedling colour. This system has great practical potential for hybrid seed breeding and production as it overcomes the problems intrinsic to other male-sterility systems and can be easily adapted for a range of tomato cultivars and diverse vegetable crops.

13.
Mol Plant ; 13(1): 42-58, 2020 Jan 06.
Artigo em Inglês | MEDLINE | ID: mdl-31678614

RESUMO

Dietary anthocyanins are important health-promoting antioxidants that make a major contribution to the quality of fruits. It is intriguing that most tomato cultivars do not produce anthocyanins in fruit. However, the purple tomato variety Indigo Rose, which has the dominant Aft locus combined with the recessive atv locus from wild tomato species, exhibits light-dependent anthocyanin accumulation in the fruit skin. Here, we report that Aft encodes a functional anthocyanin activator named SlAN2-like, while atv encodes a nonfunctional version of the anthocyanin repressor SlMYBATV. The expression of SlAN2-like is responsive to light, and the functional SlAN2-like can activate the expression of both anthocyanin biosynthetic genes and their regulatory genes, suggesting that SlAN2-like acts as a master regulator in the activation of anthocyanin biosynthesis. We further showed that cultivated tomatoes contain nonfunctional alleles of SlAN2-like and therefore fail to produce anthocyanins. Consistently, expression of a functional SlAN2-like gene driven by the fruit-specific promoter in a tomato cultivar led to the activation of the entire anthocyanin biosynthesis pathway and high-level accumulation of anthocyanins in both the peel and flesh. Taken together, our study exemplifies that efficient engineering of complex metabolic pathways could be achieved through tissue-specific expression of master transcriptional regulators.

14.
Adv Healthc Mater ; 9(2): e1901176, 2020 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-31840437

RESUMO

The development of a universal influenza vaccine is an ideal strategy to eliminate public health threats from influenza epidemics and pandemics. This ultimate goal is restricted by the low immunogenicity of conserved influenza epitopes. Layered protein nanoparticles composed of well-designed conserved influenza structures have shown improved immunogenicity with new physical and biochemical features. Herein, structure-stabilized influenza matrix protein 2 ectodomain (M2e) and M2e-neuraminidase fusion (M2e-NA) recombinant proteins are generated and M2e protein nanoparticles and double-layered M2e-NA protein nanoparticles are produced by ethanol desolvation and chemical crosslinking. Immunizations with these protein nanoparticles induce immune protection against different viruses of homologous and heterosubtypic NA in mice. Double-layered M2e-NA protein nanoparticles induce higher levels of humoral and cellular responses compared with their comprising protein mixture or M2e nanoparticles. Strong cytotoxic T cell responses are induced in the layered M2e-NA protein nanoparticle groups. Antibody responses contribute to the heterosubtypic NA immune protection. The protective immunity is long lasting. These results demonstrate that double-layered protein nanoparticles containing structure-stabilized M2e and NA can be developed into a universal influenza vaccine or a synergistic component of such vaccines. Layered protein nanoparticles can be a general vaccine platform for different pathogens.

15.
Nucleic Acids Res ; 48(D1): D871-D881, 2020 Jan 08.
Artigo em Inglês | MEDLINE | ID: mdl-31665429

RESUMO

Drug combinations have demonstrated high efficacy and low adverse side effects compared to single drug administration in cancer therapies and thus have drawn intensive attention from researchers and pharmaceutical enterprises. Due to the rapid development of high-throughput screening (HTS), the number of drug combination datasets available has increased tremendously in recent years. Therefore, there is an urgent need for a comprehensive database that is crucial to both experimental and computational screening of synergistic drug combinations. In this paper, we present DrugCombDB, a comprehensive database devoted to the curation of drug combinations from various data sources: (i) HTS assays of drug combinations; (ii) manual curations from the literature; and (iii) FDA Orange Book and external databases. Specifically, DrugCombDB includes 448 555 drug combinations derived from HTS assays, covering 2887 unique drugs and 124 human cancer cell lines. In particular, DrugCombDB has more than 6000 000 quantitative dose responses from which we computed multiple synergy scores to determine the overall synergistic or antagonistic effects of drug combinations. In addition to the combinations extracted from existing databases, we manually curated 457 drug combinations from thousands of PubMed publications. To benefit the further experimental validation and development of computational models, multiple datasets that are ready to train prediction models for classification and regression analysis were constructed and other significant related data were gathered. A website with a user-friendly graphical visualization has been developed for users to access the wealth of data and download prebuilt datasets. Our database is available at http://drugcombdb.denglab.org/.

16.
J Econ Entomol ; 113(1): 399-406, 2020 Feb 08.
Artigo em Inglês | MEDLINE | ID: mdl-31756251

RESUMO

Laodelphax striatellus (Fallén) is an important rice pest species which has developed high resistance to imidacloprid. Previous studies have demonstrated that CYP6AY3v2 and CYP353D1v2 were constitutively overexpressed in a imidacloprid resistant strain and can metabolize imidacloprid to mediated metabolic resistance. Further studies still needed to explore whether there are other L. striatellus P450 enzymes that can also metabolize imidacloprid. In this study, the expression level of L. striatellus CYP4C71 was significantly upregulated both in laboratory strains and field strains of L. striatellus after imidacloprid treatment for 4 h. The capability of CYP4C71 to metabolize imidacloprid was investigated. The full-length CYP4C71 was cloned, and its open reading frame was 1,515 bp with an enzyme estimated to be 505 amino acid residues in size. Furthermore, CYP4C71 was heterologously expressed along with L. striatellus cytochrome P450 reductase (CPR) in insect cells. A carbon monoxide difference spectra analysis confirmed the successful expression of CYP4C71. The recombinant CYP4C71 showed high P450 O-demethylation activity with PNP as a substrate. In vitro metabolism studies showed that recombinant CYP4C71 can metabolize imidacloprid to an easily excreted hydroxy-form. The rate of imidacloprid depletion in response to imidacloprid concentration revealed Michaelis Menten kinetics (R2 fitted curve = 0.99) with a relative low affinity: Kcat = 0.032 ± 0.009 pmol depleted imidacloprid/min/pmol P450 and Km=85.19 ± 2.93 µM. A relative big Km (85.19 ± 2.93 µM) indicated relative low imidacloprid's affinity for the CYP4C71 enzyme. In conclusion, CYP4C71 was another P450 enzyme that can metabolize imidacloprid with a relatively low affinity.

17.
Oncol Rep ; 43(1): 75-86, 2020 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-31746400

RESUMO

Esophageal squamous cell carcinoma (ESCC) is a common type of esophageal cancer and is prevalent worldwide. Understanding the mechanisms underlying its formation and the search for more effective therapeutic strategies are critical due to the occurrence of chemotherapeutic drug resistance. The aim of the present study was to determine the functional relevance and therapeutic potential of carbohydrate sulfotransferase 15 (CHST15) in ESCC. CHST15 levels were measured in different ESCC cell lines and evaluated in ESCC tissues using tissue chip immunohistochemistry. Cell growth and apoptosis assays, 3­(4,5­dimethylthiazol­2­yl)­2,5­diphenyltetrazolium bromide assays, and clonogenic assays were conducted using TE­1 cells and lenti­shCHST15 virus constructs were used to investigate the function of CHST15 in cell proliferation and apoptosis. mRNA microarray analysis was performed to determine the underlying mechanism of CHST15 regulation in TE­1 cell proliferation and apoptosis. The results showed that knockdown of CHST15 inhibited TE­1 cell growth and proliferation, but induced cell apoptosis. CHST15 was more frequently detected in ESCC tissue compared with that in normal esophageal tissue. Microarray data analysis indicated that the inhibition of cell proliferation and activation of cell apoptosis in CHST15­knockdown cells may be caused by altered CHST15/ILKAP/CCND1 and CHST15/RABL6/PMAIP1 signaling axes, respectively.

18.
Plant Cell ; 32(2): 429-448, 2020 Feb.
Artigo em Inglês | MEDLINE | ID: mdl-31852773

RESUMO

JASMONATE ZIM-DOMAIN (JAZ) transcriptional repressors are key regulators of jasmonate (JA) signaling in plants. At the resting stage, the C-terminal Jas motifs of JAZ proteins bind the transcription factor MYC2 to repress JA signaling. Upon hormone elicitation, the Jas motif binds the hormone receptor CORONATINE INSENSITIVE1, which mediates proteasomal degradation of JAZs and thereby allowing the Mediator subunit MED25 to activate MYC2. Subsequently, plants desensitize JA signaling by feedback generation of dominant JAZ splice variants that repress MYC2. Here we report the mechanistic function of Arabidopsis (Arabidopsis thaliana) MED25 in regulating the alternative splicing of JAZ genes through recruiting the splicing factors PRE-mRNA-PROCESSING PROTEIN 39a (PRP39a) and PRP40a. We demonstrate that JA-induced generation of JAZ splice variants depends on MED25 and that MED25 recruits PRP39a and PRP40a to promote the full splicing of JAZ genes. Therefore, MED25 forms a module with PRP39a and PRP40a to prevent excessive desensitization of JA signaling mediated by JAZ splice variants.

19.
Neural Netw ; 121: 294-307, 2020 Jan.
Artigo em Inglês | MEDLINE | ID: mdl-31586857

RESUMO

Artificial neural networks (ANNs), a popular path towards artificial intelligence, have experienced remarkable success via mature models, various benchmarks, open-source datasets, and powerful computing platforms. Spiking neural networks (SNNs), a category of promising models to mimic the neuronal dynamics of the brain, have gained much attention for brain inspired computing and been widely deployed on neuromorphic devices. However, for a long time, there are ongoing debates and skepticisms about the value of SNNs in practical applications. Except for the low power attribute benefit from the spike-driven processing, SNNs usually perform worse than ANNs especially in terms of the application accuracy. Recently, researchers attempt to address this issue by borrowing learning methodologies from ANNs, such as backpropagation, to train high-accuracy SNN models. The rapid progress in this domain continuously produces amazing results with ever-increasing network size, whose growing path seems similar to the development of deep learning. Although these ways endow SNNs the capability to approach the accuracy of ANNs, the natural superiorities of SNNs and the way to outperform ANNs are potentially lost due to the use of ANN-oriented workloads and simplistic evaluation metrics. In this paper, we take the visual recognition task as a case study to answer the questions of "what workloads are ideal for SNNs and how to evaluate SNNs makes sense". We design a series of contrast tests using different types of datasets (ANN-oriented and SNN-oriented), diverse processing models, signal conversion methods, and learning algorithms. We propose comprehensive metrics on the application accuracy and the cost of memory & compute to evaluate these models, and conduct extensive experiments. We evidence the fact that on ANN-oriented workloads, SNNs fail to beat their ANN counterparts; while on SNN-oriented workloads, SNNs can fully perform better. We further demonstrate that in SNNs there exists a trade-off between the application accuracy and the execution cost, which will be affected by the simulation time window and firing threshold. Based on these abundant analyses, we recommend the most suitable model for each scenario. To the best of our knowledge, this is the first work using systematical comparisons to explicitly reveal that the straightforward workload porting from ANNs to SNNs is unwise although many works are doing so and a comprehensive evaluation indeed matters. Finally, we highlight the urgent need to build a benchmarking framework for SNNs with broader tasks, datasets, and metrics.


Assuntos
Potenciais de Ação/fisiologia , Inteligência Artificial , Reconhecimento Automatizado de Padrão/métodos , Algoritmos , Encéfalo/fisiologia , Humanos , Memória/fisiologia , Neurônios/fisiologia
20.
Int J Mol Sci ; 20(23)2019 Nov 30.
Artigo em Inglês | MEDLINE | ID: mdl-31801264

RESUMO

MicroRNAs (miRNAs) are a highly abundant collection of functional non-coding RNAs involved in cellular regulation and various complex human diseases. Although a large number of miRNAs have been identified, most of their physiological functions remain unknown. Computational methods play a vital role in exploring the potential functions of miRNAs. Here, we present DeepMiR2GO, a tool for integrating miRNAs, proteins and diseases, to predict the gene ontology (GO) functions based on multiple deep neuro-symbolic models. DeepMiR2GO starts by integrating the miRNA co-expression network, protein-protein interaction (PPI) network, disease phenotype similarity network, and interactions or associations among them into a global heterogeneous network. Then, it employs an efficient graph embedding strategy to learn potential network representations of the global heterogeneous network as the topological features. Finally, a deep multi-label classification network based on multiple neuro-symbolic models is built and used to annotate the GO terms of miRNAs. The predicted results demonstrate that DeepMiR2GO performs significantly better than other state-of-the-art approaches in terms of precision, recall, and maximum F-measure.

SELEÇÃO DE REFERÊNCIAS
DETALHE DA PESQUISA