Structure of a pantothenate transporter and implications for ECF module sharing and energy coupling of group II ECF transporters.

Energy-coupling factor (ECF) transporters are a unique group of ATP-binding cassette (ABC) transporters responsible for micronutrient uptake from the environment. Each ECF transporter is composed of an S component (or EcfS protein) and T/A/A' components (or EcfT/A/A' proteins; ECF module). Among the group II ECF transporters, several EcfS proteins share one ECF module; however, the underlying mechanism remains unknown. Here we report the structure of a group II ECF transporter-pantothenate transporter from Lactobacillus brevis (LbECF-PanT), which shares the ECF module with the folate and hydroxymethylpyrimidine transporters (LbECF-FolT and LbECF-HmpT). Structural and mutational analyses revealed the residues constituting the pantothenate-binding pocket. We found that although the three EcfS proteins PanT, FolT, and HmpT are dissimilar in sequence, they share a common surface area composed of the transmembrane helices 1/2/6 (SM1/2/6) to interact with the coupling helices 2/3 (CH2/3) of the same EcfT. CH2 interacts mainly with SM1 via hydrophobic interactions, which may modulate the sliding movement of EcfS. CH3 binds to a hydrophobic surface groove formed by SM1, SM2, and SM6, which may transmit the conformational changes from EcfA/A' to EcfS. We also found that the residues at the intermolecular surfaces in LbECF-PanT are essential for transporter activity, and that these residues may mediate intermolecular conformational transmission and/or affect transporter complex stability. In addition, we found that the structure of EcfT is conformationally dynamic, which supports its function as a scaffold to mediate the interaction of the ECF module with various EcfS proteins to form different transporter complexes.

Learning Representations from Heart Sound: A Comparative Study on Shallow and Deep Models.

Leveraging the power of artificial intelligence to facilitate an automatic analysis and monitoring of heart sounds has increasingly attracted tremendous efforts in the past decade. Nevertheless, lacking on standard open-access database made it difficult to maintain a sustainable and comparable research before the first release of the PhysioNet CinC Challenge Dataset. However, inconsistent standards on data collection, annotation, and partition are still restraining a fair and efficient comparison between different works. To this line, we introduced and benchmarked a first version of the Heart Sounds Shenzhen (HSS) corpus. Motivated and inspired by the previous works based on HSS, we redefined the tasks and make a comprehensive investigation on shallow and deep models in this study. First, we segmented the heart sound recording into shorter recordings (10 s), which makes it more similar to the human auscultation case. Second, we redefined the classification tasks. Besides using the 3 class categories (normal, moderate, and mild/severe) adopted in HSS, we added a binary classification task in this study, i.e., normal and abnormal. In this work, we provided detailed benchmarks based on both the classic machine learning and the state-of-the-art deep learning technologies, which are reproducible by using open-source toolkits. Last but not least, we analyzed the feature contributions of best performance achieved by the benchmark to make the results more convincing and interpretable.

Chemical reduction of three-dimensional silica micro-assemblies into microporous silicon replicas.

The carbothermal reduction of silica into silicon requires the use of temperatures well above the silicon melting point (> or =2,000 degrees C). Solid silicon has recently been generated directly from silica at much lower temperatures (< or =850 degrees C) via electrochemical reduction in molten salts. However, the silicon products of such electrochemical reduction did not retain the microscale morphology of the starting silica reactants. Here we demonstrate a low-temperature (650 degrees C) magnesiothermic reduction process for converting three-dimensional nanostructured silica micro-assemblies into microporous nanocrystalline silicon replicas. The intricate nanostructured silica microshells (frustules) of diatoms (unicellular algae) were converted into co-continuous, nanocrystalline mixtures of silicon and magnesia by reaction with magnesium gas. Selective magnesia dissolution then yielded an interconnected network of silicon nanocrystals that retained the starting three-dimensional frustule morphology. The silicon replicas possessed a high specific surface area (>500 m(2) g(-1)), and contained a significant population of micropores (< or =20 A). The silicon replicas were photoluminescent, and exhibited rapid changes in impedance upon exposure to gaseous nitric oxide (suggesting a possible application in microscale gas sensing). This process enables the syntheses of microporous nanocrystalline silicon micro-assemblies with multifarious three-dimensional shapes inherited from biological or synthetic silica templates for sensor, electronic, optical or biomedical applications.

Somatisation Disorder Detection via Speech: Introducing a Self-Supervised Learning Model.

With the depressive psychiatric disorders becoming more common, people are gradually starting to take it seriously. Somatisation disorders, as a general mental disorder, are rarely accurately identified in clinical diagnosis for its specific nature. In the previous work, speech recognition technology has been successfully applied to the task of identifying somatisation disorders on the Shenzhen Somatisation Speech Corpus. Nevertheless, there is still a scarcity of labels for somatisation disorder speech database. The current mainstream approaches in the speech recognition heavily rely on the well labelled data. Compared to supervised learning, self-supervised learning is able to achieve the same or even better recognition results while reducing the reliance on labelled samples. Moreover, self-supervised learning can generate general representations without the need for human hand-crafted features depending on the different recognition tasks. To this end, we apply self-supervised learning pre-trained models to solve few-labelled somatisation disorder speech recognition. In this study, we compare and analyse the results of three self-supervised learning models (contrastive predictive coding, wav2vec and wav2vec 2.0). The best result of wav2vec 2.0 model achieves 77.0 % unweighted average recall and is significantly better than CPC (p < .005), performing better than the benchmark of the supervised learning model.Clinical relevance- This work proposed a self-supervised learning model to resolve the few-labelled SD speech data, which can be well used for helping psychiatrists with clinical assistant to diagnosis. With this model, psychiatrists no longer need to spend a lot of time labelling SD speech data.

Effect of Ta Interlayers on Texture and Magnetic Properties of FeSi Films with Micrometer Thickness.

Magnetized soft ferromagnetic films with micrometer thickness were studied. A FeSi film, with a total thickness of 2000 nm, separated by 10 nm-thick Ta interlayers, was fabricated using the direct-current magnetron sputtering technique. The thickness of each FeSi layer between adjacent Ta layers was 100 nm. Hysteresis loop measurement was used to characterize the magnetic properties of the layer. X-ray diffraction patterns and high-resolution transmission electron microscopy were used to characterize its texture. The experimental results showed that the FeSi film separated by Ta interlayers exhibited a lower saturation magnetization and a higher coercivity than those of the 1140 nm-thick FeSi film. The insertion of Ta interlayers resulted in the disappearance of the crystal plane of FeSi (221), and better texture of the crystal plane of FeSi (210). The FeSi film exhibited a crystal plane of FeSi (210) with a bcc crystalline structure. The Ta interlayers were partially amorphous, exhibiting crystal plane of Ta (002) and TaSi2 (310). The matching of magnetic properties between interlayers and soft magnetic layers played an important role in maintaining its soft magnetic properties.

A Federated Learning Paradigm for Heart Sound Classification.

Cardiovascular diseases (CVDs) have been ranked as the leading cause for deaths. The early diagnosis of CVDs is a crucial task in the medical practice. A plethora of efforts were given to the automated auscultation of heart sound, which leverages the power of computer audition to develop a cheap, non-invasive method that can be used at any time and anywhere for measuring the status of the heart. Nevertheless, previous works ignore an important factor, namely, the privacy of the user data. On the one hand, learnt models are always hungry for bigger data. On the other hand, it can be difficult to protect personal private information when collecting such large amount of data. In this dilemma, we propose a federated learning (FL) framework for the heart sound classification task. To the best of our knowledge, this is the first time to introduce FL to this field. We conducted multiple experiments, analysed the impact of data distribution across collaborative institutions on model quality and learning patterns, and verified the feasibility and effectiveness of FL based on real data. Non- independent identically distributed (Non-IID) data and model quality can be effectively improved by adding a strategy of globally sharing data.

Magnetic double-core@shell MnO2@NiFe@DE as a multifunctional scavenger for efficient removal of tetracycline, anionic and cationic dyes.

Design and fabrication of core-shell nanomaterials with excellent properties such as multifunctionality, tunability, and stability for the removal of recalcitrant pollutants from wastewater is highly valued. In this work, magnetic MnO2@NiFe@DE nanocomposites with double-core@shell structures were obtained via a two-step hydrothermal method for efficiently removing tetracycline, anionic and cationic dyes through the synergistic effect of oxidation and adsorption. The novel nanomaterial displayed superior removal of methyl orange, methylene blue, and tetracycline in low pH solutions with 100%, 100%, and 83%, respectively. The effects of solution pH, adsorption time, and contaminant concentration on the performance of the nanocomposite were also investigated, and the pseudo-second-order kinetic model well described the data. Physical adsorption including electrostatic adsorption, anion exchange, and hydrogen bonding are the predominant mechanisms for contaminant removal. The oxidation mechanism is mainly hydroxyl radical action. Through the use of permanent magnets, the recovery process of the adsorbent and the adsorbed dyes and antibiotics is energetically and economically sustainable. This as-synthesized nanocomposite as multifunction material has a high removal rate, low cost, and easy separation, and the applicability in treating the solutions with low pH, which is promised to be an efficient organic wastewater remover in practical applications.

Using Flame-Assisted Printing to Fabricate Large Nanostructured Oxide Thin Film for Electrochromic Applications.

Flame spray pyrolysis was a process to produce oxide nanoparticles in a self-sustaining flame. When the produced nanoparticles were deposited on a substrate, nanostructured oxide thin films could be obtained. However, the size of the thin film was usually limited by the fixed substrate. Here, we demonstrated that thin film with a large area could be deposited by using the moving substrate, which was precisely controlled by servo motors. As a result, the flame tip could scan over the substrate and deposit the nanoparticles on it line by line, analogues to a printing process called flame-assisted printing (FAP). As an example, nanostructured bismuth-oxide thin films with a size of up to 20 cm × 20 cm were deposited with the FAP process. The bismuth-oxide thin film exhibited a stable electrochromic property with a high modulation of 70.5%. The excellent performance could be ascribed to its porous nanostructure formed in the FAP process. The process can be extended to deposit other various oxides (e.g., tungsten-oxide) thin films with a large size for versatile applications.

Synthesis of Metal Oxide/Carbon Nanofibers via Biostructure Confinement as High-Capacity Anode Materials.

For applications in energy storage and conversion, many metal oxide (MO)/C composite fibers have been synthesized using cellulose as the template. However, MO particles in carbon fibers usually experience anomalous growth to a size of >200 nm, which is detrimental to the overall performance of the composite. In this paper, we report the successful development of a generic approach to synthesize a fiber composite with highly dispersed MO nanoparticles (10-80 nm) via simple swelling, nitrogen doping, and carbonization of the cellulose microfibril. The growth of the MO nanoparticles is confined by the structure of the microfibrils. Density functional theory calculation further reveals that the doped N atoms supply ample nucleation sites for size confinement of the nanoparticles. The encapsulation structure of small MO nanoparticles in the conductive carbon matrix improves their electrochemical performance. For example, the formed SnOx/carbon nanocomposite exhibits high specific capacities of 1011.0 mA h g-1 at 0.5 A g-1 and 581.8 mA h g-1 at 5 A g-1. Moreover, the fiber-like nanocomposite can be combined with carbon nanotubes to form a flexible binder-free electrode with a capacity of ∼10 mA h cm-2, far beyond the commercial level. The process developed in this study offers an alternative approach to sophisticated electrospinning for the synthesis of other fiber-like MO/carbon nanocomposites for versatile applications.

Synthesis of Porous Si/C Composite Nanosheets from Vermiculite with a Hierarchical Structure as a High-Performance Anode for Lithium-Ion Battery.

Silicon nanosheets are fascinating anode materials for lithium-ion batteries because of their high specific capacities, structural stability, and fast kinetics in alloying/dealloying with Li. The nanosheets can be synthesized through chemical vapor deposition (CVD), topochemical reaction, and templating method. After coating with a carbon nanolayer, they exhibit enhanced electrochemical performance. However, it is challenging to synthesize ultrathin carbon-coated silicon nanosheets. In this work, porous silicon/carbon (pSi/C) composite nanosheets are synthesized by reducing the carbon-coated expanded vermiculite with metallic Al in the molten salts. The as-prepared pSi/C nanosheets retain the layered nanostructure of vermiculite, with a thickness of less than 50 nm. The carbon nanolayer serves as the diffusion barrier and mechanical support for the growth of mesoporous silicon nanosheets. The anode of pSi/C nanosheets achieves remarkable electrochemical performance, exhibiting a reversible capacity of 1837 mA h g-1 at 4 A g-1 and retaining 71.5% of the initial capacity after 500 cycles. The process can be extended to the synthesis of the pSi/C composite nanotube by using other carbon-coated silicate templates such as halloysite.

3D Interconnected V6O13 Nanosheets Grown on Carbonized Textile via a Seed-Assisted Hydrothermal Process as High-Performance Flexible Cathodes for Lithium-Ion Batteries.

Three-dimensional (3D) free-standing nanostructured materials have been proven to be one of the most promising electrodes for energy storage due to their enhanced electrochemical performance. And they are also widely studied for the wearable energy storage systems. In this work, interconnected V6O13 nanosheets were grown on the flexible carbonized textile (c-textile) via a seed-assisted hydrothermal method to form a 3D free-standing electrode for lithium-ion batteries (LIBs). The electrode exhibited a specific capacity of 170 mA h g-1 at a specific current of 300 mA g-1. With carbon nanotube (CNT) coating, its specific capacities further increased 12-40% at the various current rates. It could retain a reversible capacity of 130 mA h g-1, 74% of the initial capacity after 300 cycles at the specific current of 300 mA g-1. It outperformed most of the mixed-valence vanadium oxides. The improved electrochemical performance was ascribed to the synergistic effect of the 3D nanostructure of V6O13 for feasible Li+ diffusion and transport and highly conductive hierarchical conductive network formed by CNT and carbon fiber in c-textile.

Adsorption of Tetracycline with Reduced Graphene Oxide Decorated with MnFe2O4 Nanoparticles.

Nanomaterials were widely used as efficient adsorbents for environmental remediation of tetracycline pollution. However, the separation of the adsorbents posed the challenge to their practical applications. In this study, we grew magnetic MnFe2O4 nanoparticles on the reduced graphene oxide (rGO) to form MnFe2O4/rGO nanocomposite with a one-step method. When used as the absorbent of Tetracycline, it exhibited an adsorption capacity of 41 mg/g. The adsorption kinetics and isotherm were fitted well with the pseudo-second order model and Freundlich model, respectively. The MnFe2O4/rGO nanocomposite could be easily extracted from the solution with the external magnetic field and regenerated with acid washing.

Flexible and Hierarchically Structured Sulfur Composite Cathode Based on the Carbonized Textile for High-Performance Li-S Batteries.

Carbon hosts have been utilized to obtain composite cathodes with high sulfur loadings for Li-S batteries. However, the complicated synthesis process may hinder their practical applications. Their mechanical and electrochemical properties shall be further improved. Herein, a facile scalable dip-coating process is developed to synthesize a flexible composite cathode with a high sulfur loading. Via the process, a hybrid composed of carbon nanotubes, carbon black, sulfur, and titania nanoparticles is successfully conformally coated on the carbonized textile (c-textile). The formed flexible c-textile@S/TiO2 cathodes with sulfur loadings of 1.5 and 3.0 mg cm-2 can deliver reversible discharge capacities of 860 and 659 mA h g-1 at 2 C, respectively. For the latter one, it can retain 94% of the initial capacity after 400 cycles with a high Coulombic efficiency (∼96%). When its sulfur loading is further increased to 7.0 mg cm-2, its areal capacity reaches 5.2 mA h cm-2. Such excellent performance is ascribed to the synergy effect of the three-dimension conductive hierarchical pore structure and TiO2 additive. They can physically and chemically entrap the soluble polysulfides in the composite cathode. The as-synthesized free-standing composite electrode is of low cost and a high areal capacity, making it suitable for flexible energy storage applications based on Li-S batteries.

Formation of Si Hollow Structures as Promising Anode Materials through Reduction of Silica in AlCl3-NaCl Molten Salt.

Hollow nanostructures are attractive for energy storage and conversion, drug delivery, and catalysis applications. Although these hollow nanostructures of compounds can be generated through the processes involving the well-established Kirkendall effect or ion exchange method, a similar process for the synthesis of the pure-substance one ( e. g., Si) remains elusive. Inspired by the above two methods, we introduce a continuous ultrathin carbon layer on the silica nano/microstructures (Stöber spheres, diatom frustules, sphere in sphere) as the stable reaction interface. With the layer as the diffusion mediator of the reactants, silica structures are successfully reduced into their porous silicon hollow counterparts with metal Al powder in AlCl3-NaCl molten salt. The structures are composed of silicon nanocrystallites with sizes of 15-25 nm. The formation mechanism can be explained as an etching-reduction/nucleation-growth process. When used as the anode material, the silicon hollow structure from diatom frustules delivers specific capacities of 2179, 1988, 1798, 1505, 1240, and 974 mA h g-1 at 0.5, 1, 2, 4, 6, and 8 A g-1, respectively. After being prelithiated, it retains 80% of the initial capacity after 1100 cycles at 8 A g-1. This work provides a general way to synthesize versatile silicon hollow structures for high-performance lithium ion batteries due to the existence of ample silica reactants and can be extended to the synthesis of hollow structures of other materials.

Structure and mechanism of a group-I cobalt energy coupling factor transporter.

Energy-coupling factor (ECF) transporters are a large family of ATP-binding cassette transporters recently identified in microorganisms. Responsible for micronutrient uptake from the environment, ECF transporters are modular transporters composed of a membrane substrate-binding component EcfS and an ECF module consisting of an integral membrane scaffold component EcfT and two cytoplasmic ATP binding/hydrolysis components EcfA/A'. ECF transporters are classified into groups I and II. Currently, the molecular understanding of group-I ECF transporters is very limited, partly due to a lack of transporter complex structural information. Here, we present structures and structure-based analyses of the group-I cobalt ECF transporter CbiMNQO, whose constituting subunits CbiM/CbiN, CbiQ, and CbiO correspond to the EcfS, EcfT, and EcfA components of group-II ECF transporters, respectively. Through reconstitution of different CbiMNQO subunits and determination of related ATPase and transporter activities, the substrate-binding subunit CbiM was found to stimulate CbiQO's basal ATPase activity. The structure of CbiMQO complex was determined in its inward-open conformation and that of CbiO in ß, γ-methyleneadenosine 5'-triphosphate-bound closed conformation. Structure-based analyses revealed interactions between different components, substrate-gating function of the L1 loop of CbiM, and conformational changes of CbiO induced by ATP binding and product release within the CbiMNQO transporter complex. These findings enabled us to propose a working model of the CbiMNQO transporter, in which the transport process requires the rotation or toppling of both CbiQ and CbiM, and CbiN might function in coupling conformational changes between CbiQ and CbiM.

Synthesis of Mn3O 4-Based Aerogels and Their Lithium-Storage Abilities.

Mn3O4 aerogels and their graphene nanosheet (GN) composite aerogels were synthesized by a simple supercritical-ethanol process. In the process, supercritical ethanol acted as a reductant to reduce graphene oxide and MnO2 gels simultaneously. The synthesized aerogels consisted of 10-20 nm Mn3O4 nanocrystallites, with BET-specific surface areas around 60 m(2)/g. The performance of the aerogels as anode materials for lithium-ion batteries was also evaluated in this study. The results showed that Mn3O4 aerogels as anode materials exhibited a reversible capacity of 498.7 mAh/g after 60 charge/discharge cycles while the reversible capacity for Mn3O4/GN composite aerogels could further increase to 665 mAh/g. The mechanisms for the enhanced capacity retention could be attributed to their porous structures and improved electronic contact with GN addition. The process should also offer an effective and facile method to fabricate many other porous metal oxide/GN nanocomposites for low-cost, high-capacity, environmentally benign material for lithium-ion batteries.

Structures of FolT in substrate-bound and substrate-released conformations reveal a gating mechanism for ECF transporters.

Energy-coupling factor (ECF) transporters are a new family of ABC transporters that consist of four subunits, two cytoplasmic ATPases EcfA and EcfA' and two transmembrane proteins namely EcfS for substrate-specific binding and EcfT for energy coupling. Here, we report the 3.2-Å resolution crystal structure of the EcfS protein of a folate ECF transporter from Enterococcus faecalis-EfFolT, a close homologue of FolT from Lactobacillus brevis-LbFolT. Structural and biochemical analyses reveal the residues constituting the folate-binding pocket and determining the substrate-binding specificity. Structural comparison of the folate-bound EfFolT with the folate-free LbFolT contained in the holotransporter complex discloses significant conformational change at the L1 loop, and reveals a gating mechanism of ECF transporters in which the L1 loop of EcfS acts as a gate in the substrate binding and release.

Thiol-functionalized magnetite/graphene oxide hybrid as a reusable adsorbent for Hg2+ removal.

A thiol-functionalized magnetite/graphene oxide (MGO) hybrid as an adsorbent of Hg2+ was successfully synthesized by a two-step reaction. It exhibited a higher adsorption capacity compared to the bare graphene oxide and MGO due to the combined adsorption of thiol groups and magnetite nanocrystals. Its capacity reached 289.9 mg g-1 in a solution with an initial Hg2+ concentration of 100 mg l-1. After being exchanged with H+, the adsorbent could be reused. The adsorption of Hg2+ by the thiol-functionalized MGO fits well with the Freundlich isotherm model and followed pseudo-second-order kinetics.

A magnesiothermic reaction process for the scalable production of mesoporous silicon for rechargeable lithium batteries.

Mesoporous, 3-D, nanocrystalline Si has been synthesized via the magnesiothermic reduction of SiO particles at a peak temperature of only 500 °C in a scalable flow-through reactor setup. Such 3-D porous Si as an anode material exhibited high, reversible capacities (i.e., >900 mA h g(-1) after 160 charge-discharge cycles at 1000 mA g(-1)).