Deep learning-based multimodal fusion of the surface ECG and clinical features in prediction of atrial fibrillation recurrence following catheter ablation.

BACKGROUND: Despite improvement in treatment strategies for atrial fibrillation (AF), a significant proportion of patients still experience recurrence after ablation. This study aims to propose a novel algorithm based on Transformer using surface electrocardiogram (ECG) signals and clinical features can predict AF recurrence. METHODS: Between October 2018 to December 2021, patients who underwent index radiofrequency ablation for AF with at least one standard 10-second surface ECG during sinus rhythm were enrolled. An end-to-end deep learning framework based on Transformer and a fusion module was used to predict AF recurrence using ECG and clinical features. Model performance was evaluated using areas under the receiver operating characteristic curve (AUROC), sensitivity, specificity, accuracy and F1-score. RESULTS: A total of 920 patients (median age 61 [IQR 14] years, 66.3% male) were included. After a median follow-up of 24 months, 253 patients (27.5%) experienced AF recurrence. A single deep learning enabled ECG signals identified AF recurrence with an AUROC of 0.769, sensitivity of 75.5%, specificity of 61.1%, F1 score of 55.6% and overall accuracy of 65.2%. Combining ECG signals and clinical features increased the AUROC to 0.899, sensitivity to 81.1%, specificity to 81.7%, F1 score to 71.7%, and overall accuracy to 81.5%. CONCLUSIONS: The Transformer algorithm demonstrated excellent performance in predicting AF recurrence. Integrating ECG and clinical features enhanced the models' performance and may help identify patients at low risk for AF recurrence after index ablation.

Engineering PTS-based glucose metabolism for efficient biosynthesis of bacterial cellulose by Komagataeibacter xylinus.

Bacterial cellulose (BC) is a renewable biomaterial that has attracted significant attention due to its excellent properties and wide applications. Komagataeibacter xylinus CGMCC 2955 is an important BC-producing strain. It primarily produces BC from glucose while simultaneously generating gluconic acid as a by-product, which acidifies the medium and inhibits BC synthesis. To enhance glucose uptake and BC synthesis, we reconstructed the phosphoenolpyruvate-dependent glucose phosphotransferase system (PTSGlc) and strengthened glycolysis by introducing heterologous genes, resulting in a recombinant strain (GX08PTS03; Δgcd::ptsHIcrrE. coli::ptsGE. coli::pfkAE. coli). Strain GX08PTS03 efficiently utilized glucose for BC production without accumulating gluconic acid. Subsequently, the fermentation process was systematically optimized. Under optimal conditions, strain GX08PTS03 produced 7.74 g/L of BC after 6 days of static fermentation, with a BC yield of 0.39 g/g glucose, which were 87.41 % and 77.27 % higher than those of the wild-type strain, respectively. The BC produced by strain GX08PTS03 exhibited a longer fiber diameter along with a lower porosity, significantly higher solid content, crystallinity, tensile strength, and Young's modulus. This study is novel in reporting that the engineered PTSGlc-based glucose metabolism could effectively enhance the production and properties of BC, providing a future outlook for the biopolymer industry.

The MarR family regulator RmaH mediates acid tolerance of Lactococcus lactis through regulating peptidoglycan modification genes.

Lactococcus lactis, widely used in the food fermentation industry, has developed various ways to regulate acid adaptation in the process of evolution. The investigation into how peptidoglycan (PG) senses and responds to acid stress is an expanding field. Here, we addressed the regulation of murT-gatD genes which are responsible for the amidation of PG D-Glu. We found that lactic acid stress reduced murT-gatD expression, and overexpressing these genes notably decreased acid tolerance of L. lactis NZ9000, possibly due to a reduction in PG's negative charge, facilitating the influx of extracellular protons into the cell. Subsequently, using a combination of DNA pull-down assay and electrophoretic mobility shift assay (EMSA), we identified a novel MarR family regulator, RmaH, as an activator of murT-gatD transcription. Further MEME motif prediction, EMSA verification and fluorescent protein reporter assay showed that RmaH directly bound to the DNA motif 5'-KGVAWWTTTTGCT-3' located in the upstream region of murT-gatD. Beyond the mechanistic investigation of RmaH activation of murT-gatD, this study provides new insight into how peptidoglycan modification is regulated and responds to lactic acid stress.

VASE: A High-Entropy Alloy Short-Range Order Structural Descriptor for Machine Learning.

The short-range order (SRO) structure in high-entropy alloys (HEAs) is closely associated with many properties, which can be studied through density functional theory (DFT) calculations. Atomic-scale modeling and calculations require substantial computational resources, and machine learning can provide rapid estimations of DFT results. To describe SRO information in HEAs, a new descriptor based on Voronoi Analysis and Shannon Entropy (VASE) is proposed. Based on Voronoi analysis, the Shannon entropy is introduced to directly characterize atomic spatial arrangement information except for composition and atomic interactions, which is necessary for describing the disorder atomic occupancy in HEAs. The new descriptor is used for predicting the formation energy of FeCoNiAlTiCu system based on machine learning model, which is more accurate than other descriptors (Coulomb matrices, partial radial distribution functions, and Voronoi analysis). Moreover, the model trained based on VASE descriptors exhibits the best predictive performance for unrelaxed structures (24.06 meV/atom). The introduction of Shannon entropy provides an effective representation of atomic arrangement information in HEAs, which is a powerful tool for investigating the SRO phenomena.

[Knockdown of motility-related genes of Komagataeibacter xylinus and its effect on bacterial cellulose synthesis].

Bacterial cellulose (BC) is a biopolymer synthesized by bacteria, which possess excellent characteristics such as high water holding capacity, high crystallinity, and high purity. It is widely used in food, medical, cosmetics, and functional films. Komagataeibacter xylinus is a model strain used in BC synthesis research. In bacteria, motility-related genes are associated with BC synthesis, whereas in Komagataeibacter xylinus CGMCC 2955, the functions of motility-related genes and their effects on BC synthesis are not known. To address this gap, we used the λ Red recombinant system to individually knock out motA, motB, and mot2A respectively, and constructed the knockout strains K. x-ΔmotA, K. x-ΔmotB, and K. x-Δmot2A. Additionally, both motA and motB were disrupted to construct the K. x-ΔmotAB mutant. The results demonstrated that knockout strain K. x-ΔmotAB exhibited the highest BC yield, reaching (5.05±0.26) g/L, which represented an increase of approximately 24% compared to wild-type strains. Furthermore, the BC synthesized by this strain exhibited the lowest porosity, 54.35%, and displayed superior mechanical properties with a Young's modulus of up to 5.21 GPa. As knocking out motA and motB genes in K. xylinus CGMCC 2955 did not reduce BC yield; instead, it promoted BC synthesis. Consequently, this research further deepened our understanding of the relationship between motility and BC synthesis in acetic acid bacteria. The knockouts of motA and motB genes resulted in reduced BC porosity and improved mechanical properties, provides a reference for BC synthesis and membrane structure regulation modification.

A MOF-Gel Based Separator for Suppressing Redox Mediator Shuttling in Li-O2 Batteries.

Redox mediators (RMs) are widely utilized in the electrolytes of Li-O2 batteries to catalyze the formation/decomposition of Li2O2, which significantly enhances the cycling performance and reduces the charge overpotential. However, RMs have a shuttle effect by migrating to the Li anode side and inducing Li metal degradation through a parasitic reaction. Herein, a metal-organic framework gel (MOF-gel) separator is proposed to restrain the shuttling of RMs. Compared to traditional MOF nanoparticles, MOF gels form uniform and dense films on the separators. When using Ru(acac)3 (ruthenium acetylacetonate) as an RM, the MOF-gel separator suppresses the shuttling of Ru(acac)3 toward the Li anode side and significantly enhances the performance of Li-O2 batteries. Specifically, Li-O2 batteries exhibit an ultralong cycling life (410 cycles) at a current density of 0.5 A g-1. Moreover, the batteries using the MOF-gel/celgard separator exhibit significantly improved cycling performance (increase by ≈1.6 times) at a high current density of 1.0 A g-1 and a decreased charge/discharge overpotential. This result is expected to guide future development of battery separators and the exploration of redox mediators.

AcrR1, a novel TetR/AcrR family repressor, mediates acid and antibiotic resistance and nisin biosynthesis in Lactococcus lactis F44.

Lactococcus lactis, widely used in the manufacture of dairy products, encounters various environmental stresses both in natural habitats and during industrial processes. It has evolved intricate machinery of stress sensing and defense to survive harsh stress conditions. Here, we identified a novel TetR/AcrR family transcription regulator, designated AcrR1, to be a repressor for acid and antibiotic tolerance that was derepressed in the presence of vancomycin or under acid stress. The survival rates of acrR1 deletion strain ΔAcrR1 under acid and vancomycin stresses were about 28.7-fold (pH 3.0, HCl), 8.57-fold (pH 4.0, lactic acid) and 2.73-fold (300 ng/mL vancomycin) greater than that of original strain F44. We also demonstrated that ΔAcrR1 was better able to maintain intracellular pH homeostasis and had a lower affinity to vancomycin. No evident effects of AcrR1 deletion on the growth and morphology of strain F44 were observed. Subsequently, we characterized that the transcription level of genes associated with amino acids biosynthesis, carbohydrate transport and metabolism, multidrug resistance, and DNA repair proteins significantly upregulated in ΔAcrR1 using transcriptome analysis and quantitative reverse transcription-PCR assays. Additionally, AcrR1 could repress the transcription of the nisin post-translational modification gene, nisC, leading to a 16.3% increase in nisin yield after AcrR1 deletion. Our results not only refined the knowledge of the regulatory mechanism of TetR/AcrR family regulator in L. lactis, but presented a potential strategy to enhance industrial production of nisin.

Multiplexed in-situ mutagenesis driven by a dCas12a-based dual-function base editor.

Mutagenesis driving genetic diversity is vital for understanding and engineering biological systems. However, the lack of effective methods to generate in-situ mutagenesis in multiple genomic loci combinatorially limits the study of complex biological functions. Here, we design and construct MultiduBE, a dCas12a-based multiplexed dual-function base editor, in an all-in-one plasmid for performing combinatorial in-situ mutagenesis. Two synthetic effectors, duBE-1a and duBE-2b, are created by amalgamating the functionalities of cytosine deaminase (from hAPOBEC3A or hAID*Δ ), adenine deaminase (from TadA9), and crRNA array processing (from dCas12a). Furthermore, introducing the synthetic separator Sp4 minimizes interference in the crRNA array, thereby facilitating multiplexed in-situ mutagenesis in both Escherichia coli and Bacillus subtilis. Guided by the corresponding crRNA arrays, MultiduBE is successfully employed for cell physiology reprogramming and metabolic regulation. A novel mutation conferring streptomycin resistance has been identified in B. subtilis and incorporated into the mutant strains with multiple antibiotic resistance. Moreover, surfactin and riboflavin titers of the combinatorially mutant strains improved by 42% and 15-fold, respectively, compared with the control strains with single gene mutation. Overall, MultiduBE provides a convenient and efficient way to perform multiplexed in-situ mutagenesis.

High-Rate Lithium-Selenium Batteries Boosted by a Multifunctional Janus Separator Over a Wide Temperature Range of -30 °C to 60 °C.

Lithium-selenium batteries are characterized by high volumetric capacity comparable to Li-S batteries, while ≈1025 times higher electrical conductivity of Se than S is favorable for high-rate capability. However, they also suffer from the "shuttling effect" of lithium polyselenides (LPSes) and Li dendrite growth. Herein, a multifunctional Janus separator is designed by coating hierarchical nitrogen-doped carbon nanocages (hNCNC) and AlN nanowires on two sides of commercial polypropylene (PP) separator to overcome these hindrances. At room temperature, the Li-Se batteries with the Janus separator exhibit an unprecedented high-rate capability (331 mAh g-1 at 25 C) and retain a high capacity of 408 mAh g-1 at 3 C after 500 cycles. Moreover, the high retained capacities are achieved over a wide temperature range from -30 °C to 60 °C, showing the potential application under extreme environments. The excellent performances result from the "1+1>2" synergism of suppressed LPSes shuttling by chemisorption and electrocatalysis of hNCNC on the cathode side and suppressed Li-dendrite growth by thermally conductive AlN-network on the anode side, which can be well understood by the "Bucket Effect". This Janus separator provides a general strategy to develop high-performance lithium-chalcogen (Se, S, SeS2 ) batteries.

Metabolic Engineering of Saccharomyces cerevisiae for Efficient Retinol Synthesis.

Retinol, the main active form of vitamin A, plays a role in maintaining vision, immune function, growth, and development. It also inhibits tumor growth and alleviates anemia. Here, we developed a Saccharomyces cerevisiae strain capable of high retinol production. Firstly, the de novo synthesis pathway of retinol was constructed in S. cerevisiae to realize the production of retinol. Second, through modular optimization of the metabolic network of retinol, the retinol titer was increased from 3.6 to 153.6 mg/L. Then, we used transporter engineering to regulate and promote the accumulation of the intracellular precursor retinal to improve retinol production. Subsequently, we screened and semi-rationally designed the key enzyme retinol dehydrogenase to further increase the retinol titer to 387.4 mg/L. Lastly, we performed two-phase extraction fermentation using olive oil to obtain a final shaking flask retinol titer of 1.2 g/L, the highest titer reported at the shake flask level. This study laid the foundation for the industrial production of retinol.

In Silico Prediction and Mining of Exporters for Secretory Bioproduction of Terpenoids in Saccharomyces cerevisiae.

Terpenoids are the largest class of natural products, and their bioproduction by engineered cell factories receives high attention. However, excessive intracellular accumulation is one of the bottlenecks that limit the further improvement of the yield of terpenoid products. Therefore, it is important to mine exporters to achieve the secretory production of terpenoids. This study proposed a framework for the in silico prediction and mining of terpenoid exporters in Saccharomyces cerevisiae. Through the process of "mining-docking-construction-validation", we found that Pdr5 of ATP-binding cassette (ABC) transporters and Osh3 of oxysterol-binding homology (Osh) proteins can promote squalene efflux. Squalene secretion of the strain overexpressing Pdr5 and Osh3 increased to 141.1 times that of the control strain. Besides squalene, ABC exporters also can promote the secretion of ß-carotene and retinal. Molecular dynamics simulation results revealed that before exporter conformations transitioned to the "outward-open" states, the substrates might have bound to the tunnels and prepared for rapid efflux. Overall, this study provides a terpenoid exporter prediction and mining framework that may be generally used to identify exporters of other terpenoids.

Combinatorial metabolic engineering enables the efficient production of ursolic acid and oleanolic acid in Saccharomyces cerevisiae.

Ursolic acid (UA) and oleanolic acid (OA) have been demonstrated to have promising therapeutic potential as anticancer and bacteriostasis agents. Herein, via the heterologous expression and optimization of CrAS, CrAO, and AtCPR1, the de novo syntheses of UA and OA were achieved with titers of 7.4 and 3.0 mg/L, respectively. Subsequently, metabolic flux was redirected by increasing the cytosolic acetyl-CoA level and tuning the copy numbers of ERG1 and CrAS, thereby affording 483.4 mg/L UA and 163.8 mg/L OA. Furthermore, the lipid droplet compartmentalization of CrAO and AtCPR1 alongside the strengthening of the NADPH regeneration system increased the UA and OA titers to 692.3 and 253.4 mg/L in a shake flask and to 1132.9 and 433.9 mg/L in a 3-L fermenter, which is the highest UA titer reported to date. Overall, this study provides a reference for constructing microbial cell factories that can efficiently synthesize terpenoids.

Incorporation of Non-Canonical Amino Acids into Antimicrobial Peptides: Advances, Challenges, and Perspectives.

The emergence of antimicrobial resistance is a global health concern and calls for the development of novel antibiotic agents. Antimicrobial peptides seem to be promising candidates due to their diverse sources, mechanisms of action, and physicochemical characteristics, as well as the relatively low emergence of resistance. The incorporation of noncanonical amino acids into antimicrobial peptides could effectively improve their physicochemical and pharmacological diversity. Recently, various antimicrobial peptides variants with improved or novel properties have been produced by the incorporation of single and multiple distinct noncanonical amino acids. In this review, we summarize strategies for the incorporation of noncanonical amino acids into antimicrobial peptides, as well as their features and suitabilities. Recent applications of noncanonical amino acid incorporation into antimicrobial peptides are also presented. Finally, we discuss the related challenges and prospects.

APSNet: Toward Adaptive Point Sampling for Efficient 3D Action Recognition.

Observing that it is still a challenging task to deploy 3D action recognition methods in real-world scenarios, in this work, we investigate the accuracy-efficiency trade-off for 3D action recognition. We first introduce a simple and efficient backbone network structure for 3D action recognition, in which we directly extract the geometry and motion representations from the raw point cloud videos through a set of simple operations (i.e., coordinate offset generation and mini-PoinNet). Based on the backbone network, we propose an end-to-end optimized network called adaptive point sampling network (APSNet) to achieve the accuracy-efficiency trade-off, which mainly consists of three stages: the coarse feature extraction stage, the decision making stage, and the fine feature extraction stage. In APSNet, we adaptively decide the optimal resolutions (i.e., the optimal number of points) for each pair of frames based on any input point cloud video under the given computational complexity constraint. Comprehensive experiments on multiple benchmark datasets demonstrate the effectiveness and efficiency of our newly proposed APSNet for 3D action recognition.

Genomic Analysis Based on Chromosome-Level Genome Assembly Reveals an Expansion of Terpene Biosynthesis of Azadirachta indica.

Azadirachta indica (neem), an evergreen tree of the Meliaceae family, is a source of the potent biopesticide azadirachtin. The lack of a chromosome-level assembly impedes an in-depth understanding of its genome architecture and the comparative genomic analysis of A. indica. Here, a high-quality genome assembly of A. indica was constructed using a combination of data from Illumina, PacBio, and Hi-C technology, which is the first chromosome-scale genome assembly of A. indica. Based on the length of our assembly, the genome size of A. indica is estimated to be 281 Mb anchored to 14 chromosomes (contig N50 = 6 Mb and scaffold N50 = 19 Mb). The genome assembly contained 115 Mb repetitive elements and 25,767 protein-coding genes. Evolutional analysis revealed that A. indica didn't experience any whole-genome duplication (WGD) event after the core eudicot γ event, but some genes and genome segment might likely experienced recent duplications. The secondary metabolite clusters, TPS genes, and CYP genes were also identified. Comparative genomic analysis revealed that most of the A. indica-specific TPS genes and CYP genes were located on the terpene-related clusters on chromosome 13. It is suggested that chromosome 13 may play an important role in the specific terpene biosynthesis of A. indica. The gene duplication events may be responsible for the terpene biosynthesis expansion in A. indica. The genomic dataset and genomic analysis created for A. indica will shed light on terpene biosynthesis in A. indica and facilitate comparative genomic research of the family Meliaceae.

Seeking eye protection from biomass: Carbon dot-based optical blocking films with adjustable levels of blue light blocking.

The intensity of blue light in white light emitting diodes is typically higher than that of the green and red-light components in screen displays and lighting systems. To reduce the potential harm of in white light emitting diodes to the eyes, in this paper, we have used microcrystalline cellulose to synthesize biomass-based carbon dots (Bio-CD), which not only absorb short wavelength light to produce longer wavelength emissions, but also show concentration-dependent maximum excitation and maximum emission. The Bio-CDs were mixed with polyvinyl alcohol (PVA) to produce optical blocking films (OBF) that preferentially block blue light. OBFs have good transparency and also block blue light effectively. With OBFs containing 9.9% of Bio-CDs, the film blocked 99.6% and 98.6% of 395 nm light and 450 nm light respectively, and also blocked 93.4% and 97%, respectively, of the blue light emitted by computers and mobile phone screens. OBFs containing more than 9.9% Bio-CDs block blue light more than commercially available blue light blocking glasses. By adjusting the amount of Bio-CDs in the OBFs, it is possible to produce films with different degrees of blue light blocking to meet the requirements of different applications.

Co@N-CNT/MXenes in situ grown on carbon nanotube film for multifunctional sensors and flexible supercapacitors.

The rapid development of human-machine interfaces and artificial intelligence is dependent on flexible and wearable soft devices such as sensors and energy storage systems. One of the key factors for these devices is the design of a flexible electrode with high sensitivity, fast response time, and a wide working range. Here, we report the fabrication of strain sensors and all-solid-state flexible supercapacitors using Co@N-CNT/MXenes as an electrode material. The manufactured sensor shows a high tensile range (strain up to 200%) and high stability. The resistance change caused by the fingers touching the sensor can be used to transmit the Morse code information. Flexible supercapacitors serving as power supply demonstrate excellent cycling stability (85 000 cycles) and coulombic efficiency (99.7%) for their high surface area and pseudocapacitance. A self-powered integrated system composed of the strain sensor and flexible supercapacitor is fabricated and operates stably in a wide strain sensing test range. Moreover, the flexible solar-charging self-powered integrated system could be attached to the human body for stable human motion detection. This study clearly shows that appropriate selection of a single functional material to enable it to be used in multi-functional sensors and supercapacitors can simplify the process and reduce the cost of manufacturing wearable devices.

Combinational quorum sensing devices for dynamic control in cross-feeding cocultivation.

Quorum sensing (QS) offers cell density dependent dynamic regulations in cell culture through devices such as synchronized lysis circuit (SLC) and metabolic toggle switch (MTS). However, there is still a lack of studies on cocultivation with a combination of different QS-based devices. Taking the production of isopropanol and salidroside as case studies, we have mathematically modeled a comprehensive set of QS-regulated cocultivation schemes and constructed experimental combinations of QS devices, respectively, to evaluate their feasibility and optimality for regulating growth competition and corporative production. Glucose split ratio is proposed for the analysis of competition between cell growth and targeted production. Results show that the combination of different QS devices across multiple members offers a new tool with the potential to effectively coordinate synthetic microbial consortia for achieving high product titer in cross-feeding cocultivation. It is also evident that the performance of such systems is significantly affected by dynamic characteristics of chosen QS devices, carbon source control and the operational settings. This study offers insights for future applications of combinational QS devices in synthetic microbial consortia.

Vertical and horizontal quorum-sensing-based multicellular communications.

Quorum sensing (QS) plays an important role in both natural and synthetic microbial systems. The complexity of QS entails multilayer controls, biomolecular crosstalk, and population-based interactions. In this review, we divide complex QS-based interactions into vertical and horizontal interactions. With respect to the former, we discuss QS-based interactions among phages, bacteria, and hosts in natural microbial systems, which are based on various QS signals and hormones. With regard to the latter, we highlight manipulations of QS-based interactions for multicomponent synthetic microbial consortia. We further present the recent and emerging applications of manipulating these interactions (collectively referred to as 'QS communication networks') in natural and synthetic microbiota. Finally, we identify key challenges in engineering diverse QS communication networks for various future applications.

Combinational Antibacterial Activity of Nisin and 3-Phenyllactic Acid and Their Co-production by Engineered Lactococcus lactis.

Nisin produced by certain Lactococcus lactis strains is commercially used in meat and dairy industries because of its effective antibacterial activity and food safety characteristics. It has been proved that the antibacterial activity could be enhanced when combined with other antimicrobial agents. In this study, we demonstrated that nisin and 3-phenyllactic acid (PLA) in combination displayed excellent combinational antibacterial activity against foodborne pathogens including S. xylosus and M. luteus. The potential application in food preservation was further verified via microbial analysis during the storage of meat and milk, and determination of strawberry rot rate. Scanning electron microscopy observation indicated a distinct mode of PLA with nisin, which may target at the dividing cell, contributing to their combinational antibacterial effect of nisin and PLA. Considering the positive results, a nisin-PLA co-producing strain was constructed based on the food-grade strain L. lactis F44, a nisin Z producer. By the knockout of two L-lactate dehydrogenase (LDH) and overexpression of D-LDH Y25A, the yield of PLA was significantly increased 1.77-fold in comparison with the wild type. Anti-bacterial assays demonstrated that the fermentation product of the recombinant strain performed highly effective antibacterial activity. These results provided a promising prospect for the nisin-PLA co-expressing L. lactis in food preservation on account of its considerable antibacterial activity and cost-effective performance.