Explore the vast range of titles available.

We introduce a data-driven forecasting method for high-dimensional chaotic systems using long short-term memory (LSTM) recurrent neural networks. The proposed LSTM neural networks perform inference of high-dimensional dynamical systems in their reduced order space and are shown to be an effective set of nonlinear approximators of their attractor. We demonstrate the forecasting performance of the LSTM and compare it with Gaussian processes (GPs) in time series obtained from the Lorenz 96 system, the Kuramoto–Sivashinsky equation and a prototype climate model. The LSTM networks outperform the GPs in short-term forecasting accuracy in all applications considered. A hybrid architecture, extending the LSTM with a mean stochastic model (MSM–LSTM), is proposed to ensure convergence to the invariant measure. This novel hybrid method is fully data-driven and extends the forecasting capabilities of LSTM networks.

Froth flotation is the most effective industrial method used to separate fine-grained minerals. The main problem of complex ore flotation is the negative effect of interactions among minerals in slurry, leading to variation in surface properties during separation. In this review, studies on the interactive effect among minerals on the flotation of iron ores, magnesite ores, and scheelite ores are summarized, and the main problems and mechanisms that diminish the separation efficiency of minerals are revealed in detail. Recent research progress on the flotation of these ores has confirmed that mineral aggregation, coating, and dissolution, as well as other factors caused by interacting behavior, explain the depressing effects of fine particles on mineral separation. Solvable methods for these effects are further discussed. Novel flotation processes and more selective reagents are critical for further investigations on various approaches to improve the beneficiation efficiency of these ores. This review aims to provide a good reference for conducting studies related to complex ore flotation.

The discovery of intrinsic ferromagnetism in ultrathin two-dimensional van der Waals crystals opens up exciting prospects for exploring magnetism in the ultimate two-dimensional limit. Here, we show that environmentally stable CrSe 2 nanosheets can be readily grown on a dangling-bond-free WSe 2 substrate with systematically tunable thickness down to the monolayer limit. These CrSe 2 /WSe 2 heterostructures display high-quality van der Waals interfaces with well-resolved moiré superlattices and ferromagnetic behaviour. We find no apparent change in surface roughness or magnetic properties after months of exposure in air. Our calculations suggest that charge transfer from the WSe 2 substrate and interlayer coupling within CrSe 2 play a critical role in the magnetic order in few-layer CrSe 2 nanosheets. The highly controllable growth of environmentally stable CrSe 2 nanosheets with tunable thickness defines a robust two-dimensional magnet for fundamental studies and potential applications in magnetoelectronic and spintronic devices. CrSe 2 nanosheets grown on WSe 2 show no apparent change in surface roughness or magnetic properties after months of exposure in air. Calculations suggest that charge transfer from the WSe 2 substrate and interlayer coupling within CrSe 2 play a critical role.

Intervertebral disc degeneration (IDD) is the most common cause of low-back pain. Accumulating evidence indicates that the expression profiling of noncoding RNAs (ncRNAs), including microRNAs (miRNAs), circular RNAs (circRNAs), and long noncoding RNAs (lncRNAs), are different between intervertebral disc tissues obtained from healthy individuals and patients with IDD. However, the roles of ncRNAs in IDD are still unclear until now. In this review, we summarize the studies concerning ncRNA interactions and regulatory functions in IDD. Apoptosis, aberrant proliferation, extracellular matrix degradation, and inflammatory abnormality are tetrad fundamental pathologic phenotypes in IDD. We demonstrated that ncRNAs are playing vital roles in apoptosis, proliferation, ECM degeneration, and inflammation process of IDD. The ncRNAs participate in underlying mechanisms of IDD in different ways. MiRNAs downregulate target genes’ expression by directly binding to the 3′-untranslated region of mRNAs. CircRNAs and lncRNAs act as sponges or competing endogenous RNAs by competitively binding to miRNAs and regulating the expression of mRNAs. The lncRNAs, circRNAs, miRNAs, and mRNAs widely crosstalk and form complex regulatory networks in the degenerative processes. The current review presents novel insights into the pathogenesis of IDD and potentially sheds light on the therapeutics in the future.

Spore killers in fungi are selfish genetic elements that distort Mendelian segregation in their favor. It remains unclear how many species harbor them and how diverse their mechanisms are. Here, we discover two spore killers from a natural isolate of the fission yeast Schizosaccharomyces pombe. Both killers belong to the previously uncharacterized wtf gene family with 25 members in the reference genome. These two killers act in strain-background-independent and genome-location-independent manners to perturb the maturation of spores not inheriting them. Spores carrying one killer are protected from its killing effect but not that of the other killer. The killing and protecting activities can be uncoupled by mutation. The numbers and sequences of wtf genes vary considerably between S. pombe isolates, indicating rapid divergence. We propose that wtf genes contribute to the extensive intraspecific reproductive isolation in S. pombe, and represent ideal models for understanding how segregation-distorting elements act and evolve. During evolution, new species emerge when individuals from different populations of similar organisms no longer breed with each other, or when the offspring produced if they do breed are sterile. This process is known as “reproductive isolation” and, for over 100 years, evolutionary biologists have tried to better understand how this process happens. Animals, plants and fungi produce sex cells – known as gametes – when they are preparing to reproduce. These cells are made when cells containing two copies of every gene in the organism divide to produce new cells that each only have one copy of each gene. Therefore, a particular gene copy usually has a 50% chance of being carried by an individual gamete. There are genes that selfishly increase their chances of being transmitted to the next generation by destroying the gametes that do not carry them. These “gamete killer” genes reduce the fertility of the organism and lead to reproductive isolation. Fission yeast is a fungus that is widely used in research. There are different strains of fission yeast that are reproductively isolated from each other, but it is not known whether gamete killers are responsible for this isolation. To address this question, Hu et al. investigated the causes of reproductive isolation in fission yeast. The experiments identified two gamete killers, referred to as cw9 and cw27. Both genes belong to the wtf gene family. Each gene is believed to encode two different proteins, one that acts as a poison and one that acts as an antidote. The poison is capable of killing all gametes, but the antidote protects the cells that contain the gamete killer gene. Further experiments show that the antidote produced by one of the gamete killer genes cannot protect cells against the poison produced by the other gene. A separate study by Nuckolls et al. found that another member of the wtf family also acts as a gamete killer in fission yeast. Together, these findings shed new light on the causes of reproductive isolation, and will contribute to deeper understanding of speciation and evolution in general.

The prediction of extreme events, from avalanches and droughts to tsunamis and epidemics, depends on the formulation and analysis of relevant, complex dynamical systems. Such dynamical systems are characterized by high intrinsic dimensionality with extreme events having the form of rare transitions that are several standard deviations away from the mean. Such systems are not amenable to classical order-reduction methods through projection of the governing equations due to the large intrinsic dimensionality of the underlying attractor as well as the complexity of the transient events. Alternatively, data-driven techniques aim to quantify the dynamics of specific, critical modes by utilizing data-streams and by expanding the dimensionality of the reduced-order model using delayed coordinates. In turn, these methods have major limitations in regions of the phase space with sparse data, which is the case for extreme events. In this work, we develop a novel hybrid framework that complements an imperfect reduced order model, with data-streams that are integrated though a recurrent neural network (RNN) architecture. The reduced order model has the form of projected equations into a low-dimensional subspace that still contains important dynamical information about the system and it is expanded by a long short-term memory (LSTM) regularization. The LSTM-RNN is trained by analyzing the mismatch between the imperfect model and the data-streams, projected to the reduced-order space. The data-driven model assists the imperfect model in regions where data is available, while for locations where data is sparse the imperfect model still provides a baseline for the prediction of the system state. We assess the developed framework on two challenging prototype systems exhibiting extreme events. We show that the blended approach has improved performance compared with methods that use either data streams or the imperfect model alone. Notably the improvement is more significant in regions associated with extreme events, where data is sparse.

The kinetics and catalytic mechanism of sulfamethazine (SMT) degradation using Fe 3 O 4 /Mn 3 O 4 nanocomposite as catalysts in heterogeneous Fenton-like process were investigated. The degradation process of SMT conformed to first-order kinetic model. The apparent activation energy ( E a ) of the process was calculated to be 40.5 kJ/mol. The reusability and stability of the catalysts were evaluated based on the results of the successive batch experiments. The intermediates were identified and quantified by ion chromatography (IC), high-performance liquid chromatography (HPLC), and gas chromatography–mass spectrometry (GC-MS). The results suggested that the bonds of S–C, N–C, and S–N were broken mainly by ·OH attack to form the organic compounds, which were gradually decomposed into small-molecule organic acids, such as oxalic acid, propionic acid, and formic acid. The possible catalytic mechanism for SMT degradation was tentatively proposed.

The Monte Carlo method is the most commonly used dose calculation method in the field of boron neutron capture therapy (BNCT). General-purpose Monte Carlo (MC) code (e.g., MCNP) has been used in most treatment planning systems (TPS) to calculate dose distribution, which takes overmuch time in radiotherapy planning. Based on this, we developed COMPASS (COMpact PArticle Simulation System), an MC engine specifically for BNCT dose calculation. Several optimization algorithms are used in COMPASS to make it faster than general-purpose MC code. The parallel computation of COMPASS is performed by the message passing interface (MPI) library and OpenMP commands, which allows the user to increase computational speed by increasing the computer configurations. The physical dose of each voxel is calculated for developing a treatment plan. Comparison results show that the computed dose distribution of COMPASS is in good agreement with MCNP, and the computational efficiency is better than MCNP. These results validate that COMPASS has better performance than MCNP in BNCT dose calculation.

Two-dimensional (2D) materials 1 , 2 and the associated van der Waals (vdW) heterostructures 3 – 7 have provided great flexibility for integrating distinct atomic layers beyond the traditional limits of lattice-matching requirements, through layer-by-layer mechanical restacking or sequential synthesis. However, the 2D vdW heterostructures explored so far have been usually limited to relatively simple heterostructures with a small number of blocks 8 – 18 . The preparation of high-order vdW superlattices with larger number of alternating units is exponentially more difficult, owing to the limited yield and material damage associated with each sequential restacking or synthesis step 8 – 29 . Here we report a straightforward approach to realizing high-order vdW superlattices by rolling up vdW heterostructures. We show that a capillary-force-driven rolling-up process can be used to delaminate synthetic SnS 2 /WSe 2 vdW heterostructures from the growth substrate and produce SnS 2 /WSe 2 roll-ups with alternating monolayers of WSe 2 and SnS 2 , thus forming high-order SnS 2 /WSe 2 vdW superlattices. The formation of these superlattices modulates the electronic band structure and the dimensionality, resulting in a transition of the transport characteristics from semiconducting to metallic, from 2D to one-dimensional (1D), with an angle-dependent linear magnetoresistance. This strategy can be extended to create diverse 2D/2D vdW superlattices, more complex 2D/2D/2D vdW superlattices, and beyond-2D materials, including three-dimensional (3D) thin-film materials and 1D nanowires, to generate mixed-dimensional vdW superlattices, such as 3D/2D, 3D/2D/2D, 1D/2D and 1D/3D/2D vdW superlattices. This study demonstrates a general approach to producing high-order vdW superlattices with widely variable material compositions, dimensions, chirality and topology, and defines a rich material platform for both fundamental studies and technological applications. A simple but flexible technique based on a capillary-force-driven rolling-up process produces high-order van der Waals superlattices that are hard to produce with existing fabrication techniques.

In this paper, we develop an algorithm to solve nonlinear system of monotone equations, which is a combination of a modified spectral PRP (Polak-Ribière-Polyak) conjugate gradient method and a projection method. The search direction in this algorithm is proved to be sufficiently descent for any line search rule. A line search strategy in the literature is modified such that a better step length is more easily obtained without the difficulty of choosing an appropriate weight in the original one. Global convergence of the algorithm is proved under mild assumptions. Numerical tests and preliminary application in recovering sparse signals indicate that the developed algorithm outperforms the state-of-the-art similar algorithms available in the literature, especially for solving large-scale problems and singular ones.