A Novel Hybrid Data-Driven Model for Daily Land Surface Temperature Forecasting Using Long Short-Term Memory Neural Network Based on Ensemble Empirical Mode Decomposition.

Daily land surface temperature (LST) forecasting is of great significance for application in climate-related, agricultural, eco-environmental, or industrial studies. Hybrid data-driven prediction models using Ensemble Empirical Mode Composition (EEMD) coupled with Machine Learning (ML) algorithms are useful for achieving these purposes because they can reduce the difficulty of modeling, require less history data, are easy to develop, and are less complex than physical models. In this article, a computationally simple, less data-intensive, fast and efficient novel hybrid data-driven model called the EEMD Long Short-Term Memory (LSTM) neural network, namely EEMD-LSTM, is proposed to reduce the difficulty of modeling and to improve prediction accuracy. The daily LST data series from the Mapoling and Zhijaing stations in the Dongting Lake basin, central south China, from 1 January 2014 to 31 December 2016 is used as a case study. The EEMD is firstly employed to decompose the original daily LST data series into many Intrinsic Mode Functions (IMFs) and a single residue item. Then, the Partial Autocorrelation Function (PACF) is used to obtain the number of input data sample points for LSTM models. Next, the LSTM models are constructed to predict the decompositions. All the predicted results of the decompositions are aggregated as the final daily LST. Finally, the prediction performance of the hybrid EEMD-LSTM model is assessed in terms of the Mean Square Error (MSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), Root Mean Square Error (RMSE), Pearson Correlation Coefficient (CC) and Nash-Sutcliffe Coefficient of Efficiency (NSCE). To validate the hybrid data-driven model, the hybrid EEMD-LSTM model is compared with the Recurrent Neural Network (RNN), LSTM and Empirical Mode Decomposition (EMD) coupled with RNN, EMD-LSTM and EEMD-RNN models, and their comparison results demonstrate that the hybrid EEMD-LSTM model performs better than the other five models. The scatterplots of the predicted results of the six models versus the original daily LST data series show that the hybrid EEMD-LSTM model is superior to the other five models. It is concluded that the proposed hybrid EEMD-LSTM model in this study is a suitable tool for temperature forecasting.

Native functions of the androgen receptor are essential to pathogenesis in a Drosophila model of spinobulbar muscular atrophy.

Spinobulbar muscular atrophy (SBMA) is a neurodegenerative disease caused by expansion of a polyglutamine tract in the androgen receptor (AR). This mutation confers toxic function to AR through unknown mechanisms. Mutant AR toxicity requires binding of its hormone ligand, suggesting that pathogenesis involves ligand-induced changes in AR. However, whether toxicity is mediated by native AR function or a novel AR function is unknown. We systematically investigated events downstream of ligand-dependent AR activation in a Drosophila model of SBMA. We show that nuclear translocation of AR is necessary, but not sufficient, for toxicity and that DNA binding by AR is necessary for toxicity. Mutagenesis studies demonstrated that a functional AF-2 domain is essential for toxicity, a finding corroborated by a genetic screen that identified AF-2 interactors as dominant modifiers of degeneration. These findings indicate that SBMA pathogenesis is mediated by misappropriation of native protein function, a mechanism that may apply broadly to polyglutamine diseases.

The post-synaptic density of human postmortem brain tissues: an experimental study paradigm for neuropsychiatric illnesses.

Recent molecular genetics studies have suggested various trans-synaptic processes for pathophysiologic mechanisms of neuropsychiatric illnesses. Examination of pre- and post-synaptic scaffolds in the brains of patients would greatly aid further investigation, yet such an approach in human postmortem tissue has yet to be tested. We have examined three methods using density gradient based purification of synaptosomes followed by detergent extraction (Method 1) and the pH based differential extraction of synaptic membranes (Methods 2 and 3). All three methods separated fractions from human postmortem brains that were highly enriched in typical PSD proteins, almost to the exclusion of pre-synaptic proteins. We examined these fractions using electron microscopy (EM) and verified the integrity of the synaptic membrane and PSD fractions derived from human postmortem brain tissues. We analyzed protein composition of the PSD fractions using two dimensional liquid chromatography tandem mass spectrometry (2D LC-MS/MS) and observed known PSD proteins by mass spectrometry. Immunoprecipitation and immunoblot studies revealed that expected protein-protein interactions and certain posttranscriptional modulations were maintained in PSD fractions. Our results demonstrate that PSD fractions can be isolated from human postmortem brain tissues with a reasonable degree of integrity. This approach may foster novel postmortem brain research paradigms in which the stoichiometry and protein composition of specific microdomains are examined.

HDAC6 rescues neurodegeneration and provides an essential link between autophagy and the UPS.

A prominent feature of late-onset neurodegenerative diseases is accumulation of misfolded protein in vulnerable neurons. When levels of misfolded protein overwhelm degradative pathways, the result is cellular toxicity and neurodegeneration. Cellular mechanisms for degrading misfolded protein include the ubiquitin-proteasome system (UPS), the main non-lysosomal degradative pathway for ubiquitinated proteins, and autophagy, a lysosome-mediated degradative pathway. The UPS and autophagy have long been viewed as complementary degradation systems with no point of intersection. This view has been challenged by two observations suggesting an apparent interaction: impairment of the UPS induces autophagy in vitro, and conditional knockout of autophagy in the mouse brain leads to neurodegeneration with ubiquitin-positive pathology. It is not known whether autophagy is strictly a parallel degradation system, or whether it is a compensatory degradation system when the UPS is impaired; furthermore, if there is a compensatory interaction between these systems, the molecular link is not known. Here we show that autophagy acts as a compensatory degradation system when the UPS is impaired in Drosophila melanogaster, and that histone deacetylase 6 (HDAC6), a microtubule-associated deacetylase that interacts with polyubiquitinated proteins, is an essential mechanistic link in this compensatory interaction. We found that compensatory autophagy was induced in response to mutations affecting the proteasome and in response to UPS impairment in a fly model of the neurodegenerative disease spinobulbar muscular atrophy. Autophagy compensated for impaired UPS function in an HDAC6-dependent manner. Furthermore, expression of HDAC6 was sufficient to rescue degeneration associated with UPS dysfunction in vivo in an autophagy-dependent manner. This study suggests that impairment of autophagy (for example, associated with ageing or genetic variation) might predispose to neurodegeneration. Morover, these findings suggest that it may be possible to intervene in neurodegeneration by augmenting HDAC6 to enhance autophagy.

The multiple functions of cysteine-string protein analyzed at Drosophila nerve terminals.

The synaptic vesicle-associated cysteine-string protein (CSP) is important for synaptic transmission. Previous studies revealed multiple defects at neuromuscular junctions (NMJs) of csp null-mutant Drosophila, but whether these defects are independent of each other or mechanistically linked through J domain mediated-interactions with heat-shock cognate protein 70 (Hsc70) has not been established. To resolve this issue, we genetically dissected the individual functions of CSP by an in vivo structure/function analysis. Expression of mutant CSP lacking the J domain at csp null-mutant NMJs fully restored normal thermo-tolerance of evoked transmitter release but did not completely restore evoked release at room temperature and failed to reverse the abnormal intraterminal Ca2+ levels. This suggests that J domain-mediated functions are essential for the regulation of intraterminal Ca2+ levels but only partially required for regulating evoked release and not required for protecting evoked release against thermal stress. Hence, CSP can also act as an Hsc70-independent chaperone protecting evoked release from thermal stress. Expression of mutant CSP lacking the L domain restored neurotransmission and partially reversed the abnormal intraterminal Ca2+ levels, suggesting that the L domain is important, although not essential, for the role of CSP in regulating intraterminal Ca2+ levels. We detected no effects of csp mutations on individual presynaptic Ca2+ signals triggered by action potentials, suggesting that presynaptic Ca2+ entry is not primarily impaired. Both the J and L domains were also required for the role of CSP in synaptic growth. Together, these results suggest that CSP has several independent synaptic functions, affecting synaptic growth, evoked release, thermal protection of evoked release, and intraterminal Ca2+ levels at rest and during stimulation.

Presynaptic regulation of neurotransmission in Drosophila by the g protein-coupled receptor methuselah.

Regulation of synaptic strength is essential for neuronal information processing, but the molecular mechanisms that control changes in neuroexocytosis are only partially known. Here we show that the putative G protein-coupled receptor Methuselah (Mth) is required in the presynaptic motor neuron to acutely upregulate neurotransmitter exocytosis at larval Drosophila NMJs. Mutations in the mth gene reduce evoked neurotransmitter release by approximately 50%, and decrease synaptic area and the density of docked and clustered vesicles. Pre- but not postsynaptic expression of normal Mth restored normal release in mth mutants. Conditional expression of Mth restored normal release and normal vesicle docking and clustering but not the reduced size of synaptic sites, suggesting that Mth acutely adjusts vesicle trafficking to synaptic sites.

Variola virus immune evasion design: expression of a highly efficient inhibitor of human complement.

Variola virus, the most virulent member of the genus Orthopoxvirus, specifically infects humans and has no other animal reservoir. Variola causes the contagious disease smallpox, which has a 30-40% mortality rate. Conversely, the prototype orthopoxvirus, vaccinia, causes no disease in immunocompetent humans and was used in the global eradication of smallpox, which ended in 1977. However, the threat of smallpox persists because clandestine stockpiles of variola still exist. Although variola and vaccinia share remarkable DNA homology, the strict human tropism of variola suggests that its proteins are better suited than those of vaccinia to overcome the human immune response. Here, we demonstrate the functional advantage of a variola complement regulatory protein over that of its vaccinia homologue. Because authentic variola proteins are not available for study, we molecularly engineered and characterized the smallpox inhibitor of complement enzymes (SPICE), a homologue of a vaccinia virulence factor, vaccinia virus complement control protein (VCP). SPICE is nearly 100-fold more potent than VCP at inactivating human C3b and 6-fold more potent at inactivating C4b. SPICE is also more human complement-specific than is VCP. By inactivating complement components, SPICE serves to inhibit the formation of the C3/C5 convertases necessary for complement-mediated viral clearance. SPICE provides the first evidence that variola proteins are particularly adept at overcoming human immunity, and the decreased function of VCP suggests one reason why the vaccinia virus vaccine was associated with relatively low mortality. Disabling SPICE may be therapeutically useful if smallpox reemerges.