Attention-based vector quantisation variational autoencoder for colour-patterned fabrics defect detection

Defect detection is an essential link in the fabric production process. Due to the diversity of patterns and scarcity of defect samples for colour-patterned fabrics, reconstruction-based unsupervised deep learning algorithms have received extensive attention in the field of fabric defect detection. Among them, unsupervised reconstruction models based on variational autoencoders (VAEs) have been shown to be effective. However, there is a problem of posterior collapse in the process of modelling parametric distributions of continuous variables by VAEs. Therefore, VAE-based defect detection methods for colour-patterned fabrics usually produce ambiguous reconstruction results, thereby affecting the defect detection performance. In this article, an attention-based vector quantisation variational autoencoder (AVQ-VAE) is proposed for colour-patterned fabric defect detection. The method adopts autoregressive modelling of discrete variables to avoid the posterior collapse problem of traditional VAEs, and utilises attention mechanism to enhance the feature representation ability of the model. AVQ-VAE consists of encoder, embedding space, decoder and attention mechanism. The encoder is used to map the input image into multiple feature vectors. Vector quantisation in embedding space is used for discretisation and autoregressive modelling of feature vectors. A decoder is used to decode discrete variables into images of the same size as the original input. Furthermore, an attention mechanism is used to capture channel and spatial correlations, which help the model focus on important information by adaptively recalibrating feature maps. Experimental results on public datasets demonstrate that the proposed method is robust and effective for colour-patterned fabric defect detection.     

Development of Deep-Learning-Based Single-Molecule Localization Image Analysis

Recent developments in super-resolution fluorescence microscopic techniques (SRM) have allowed for nanoscale imaging that greatly facilitates our understanding of nanostructures. However, the performance of single-molecule localization microscopy (SMLM) is significantly restricted by the image analysis method, as the final super-resolution image is reconstructed from identified localizations through computational analysis. With recent advancements in deep learning, many researchers have employed deep learning-based algorithms to analyze SMLM image data. This review discusses recent developments in deep-learning-based SMLM image analysis, including the limitations of existing fitting algorithms and how the quality of SMLM images can be improved through deep learning. Finally, we address possible future applications of deep learning methods for SMLM imaging.     

Chromatin Dynamics during Nucleotide Excision Repair: Histones on the Move

It has been a long-standing question how DNA damage repair proceeds in a nuclear environment where DNA is packaged into chromatin. Several decades of analysis combining in vitro and in vivo studies in various model organisms ranging from yeast to human have markedly increased our understanding of the mechanisms underlying chromatin disorganization upon damage detection and re-assembly after repair. Here, we review the methods that have been developed over the years to delineate chromatin alterations in response to DNA damage by focusing on the well-characterized Nucleotide Excision Repair (NER) pathway. We also highlight how these methods have provided key mechanistic insight into histone dynamics coupled to repair in mammals, raising new issues about the maintenance of chromatin integrity. In particular, we discuss how NER factors and central players in chromatin dynamics such as histone modifiers, nucleosome remodeling factors, and histone chaperones function to mobilize histones during repair.     

Advances in De Novo Drug Design: From Conventional to Machine Learning Methods

De novo drug design is a computational approach that generates novel molecular structures from atomic building blocks with no a priori relationships. Conventional methods include structure-based and ligand-based design, which depend on the properties of the active site of a biological target or its known active binders, respectively. Artificial intelligence, including ma-chine learning, is an emerging field that has positively impacted the drug discovery process. Deep reinforcement learning is a subdivision of machine learning that combines artificial neural networks with reinforcement-learning architectures. This method has successfully been em-ployed to develop novel de novo drug design approaches using a variety of artificial networks including recurrent neural networks, convolutional neural networks, generative adversarial networks, and autoencoders. This review article summarizes advances in de novo drug design, from conventional growth algorithms to advanced machine-learning methodologies and high-lights hot topics for further development.     

How We Make DNA Origami

DNA origami has attracted substantial attention since its invention ten years ago, due to the seemingly infinite possibilities that it affords for creating customized nanoscale objects. Although the basic concept of DNA origami is easy to understand, using custom DNA origami in practical applications requires detailed know‐how for designing and producing the particles with sufficient quality and for preparing them at appropriate concentrations with the necessary degree of purity in custom environments. Such know‐how is not readily available for newcomers to the field, thus slowing down the rate at which new applications outside the field of DNA nanotechnology may emerge. To foster faster progress, we share in this article the experience in making and preparing DNA origami that we have accumulated over recent years. We discuss design solutions for creating advanced structural motifs including corners and various types of hinges that expand the design space for the more rigid multilayer DNA origami and provide guidelines for preventing undesired aggregation and on how to induce specific oligomerization of multiple DNA origami building blocks. In addition, we provide detailed protocols and discuss the expected results for five key methods that allow efficient and damage‐free preparation of DNA origami. These methods are agarose‐gel purification, filtration through molecular cut‐off membranes, PEG precipitation, size‐exclusion chromatography, and ultracentrifugation‐based sedimentation. The guide for creating advanced design motifs and the detailed protocols with their experimental characterization that we describe here should lower the barrier for researchers to accomplish the full DNA origami production workflow.     

Exploitation of Feedback in Enzyme-catalysed Reactions

Some cellular systems, such as yeast, bacteria and slime mould, display dynamic behavior including switches and rhythms driven by feedback in enzyme-catalysed reactions. The mechanisms of these processes have been well investigated and recent attention has turned to generating similar responses in synthetic biocatalytic systems, with a view to creating bioinspired analogues for applications. Here we discuss how feedback arises in the reaction mechanisms of some enzyme-catalyzed reactions in vitro, the behaviour obtained and the emerging applications. These autocatalytic reactions may provide insights into behaviour in cellular systems as well as new methods for drug delivery, sensing and repair that can be exploited in living systems.     

Recent Development of Photothermal Agents (PTAs) Based on Small Organic Molecular Dyes

Photothermal therapy (PTT) has attracted great attention due to its noninvasive and effective use against cancer. Various photothermal agents (PTAs) including organic and inorganic PTAs have been developed in the last decades. Organic PTAs based on small-molecule dyes exhibit great potential for future clinical applications considering their good biocompatibility and easy chemical modification or functionalization. In this review, we discuss the recent progress of organic PTAs based on small-molecule dyes for enhanced PTT. We summarize the strategies to improve the light penetration of PTAs, methods to enhance their photothermal conversion efficiency, how to optimize PTAs’ delivery into deep tumors, and how to resist photobleaching under repeated laser irradiation. We hope that this review can rouse the interest of researchers in the field of PTAs based on small-molecule dyes and help them to fabricate next-generation PTAs for noninvasive cancer therapy.     

Nanoscale phase change memory materials

Phase change memory materials store information through their reversible transitions between crystalline and amorphous states. For typical metal chalcogenide compounds, their phase transition properties directly impact critical memory characteristics and the manipulation of these is a major focus in the field. Here, we discuss recent work that explores the tuning of such properties by scaling the materials to nanoscale dimensions, including fabrication and synthetic strategies used to produce nanoscale phase change memory materials. The trends that emerge are relevant to understanding how such memory technologies will function as they scale to ever smaller dimensions and also suggest new approaches to designing materials for phase change applications. Finally, the challenges and opportunities raised by integrating nanoscale phase change materials into switching devices are discussed.     

Sanger Gap Sequencing for Genetic Alphabet Expansion of DNA

Genetic alphabet expansion technology, creating new replicable and functional DNA molecules with unnatural base pairs (UBPs), is the novel promising research area of xenobiology. Recently, this technology has rapidly advanced, resulting in the need for a sequencing method for DNA molecules containing UBPs. However, all of the conventional sequencing methods, such as Sanger methods, are for four-letter DNA molecules. Here, we present an improved Sanger sequencing method (Sanger gap sequencing) for DNAs containing our UBP, Ds-Px, which appears as gaps in the sequencing peak patterns. By improving the sequencing reaction for efficient Ds-Px pairing and using modified Px substrates, we have developed a sequencing method with increased processivity and clear gap patterns for multiple Ds-Px pairs in various sequence contexts. This method is useful for UBP applications such as high-affinity DNA aptamer generation and semisynthetic organism creation involving UBPs. In addition, through this research, we found that the side chains of UBs greatly affect the efficiency of UB pairings in replication, thus suggesting further development of UBPs.     

Applications of next-generation sequencing technologies to diagnostic virology

Novel DNA sequencing techniques, referred to as "next-generation" sequencing (NGS), provide high speed and throughput that can produce an enormous volume of sequences with many possible applications in research and diagnostic settings. In this article, we provide an overview of the many applications of NGS in diagnostic virology. NGS techniques have been used for high-throughput whole viral genome sequencing, such as sequencing of new influenza viruses, for detection of viral genome variability and evolution within the host, such as investigation of human immunodeficiency virus and human hepatitis C virus quasispecies, and monitoring of low-abundance antiviral drug-resistance mutations. NGS techniques have been applied to metagenomics-based strategies for the detection of unexpected disease-associated viruses and for the discovery of novel human viruses, including cancer-related viruses. Finally, the human virome in healthy and disease conditions has been described by NGS-based metagenomics.     

Polymeric lipid assemblies as novel theranostic tools

Polymerizable lipids have been used in research and medical applications such as membrane models, imaging platforms, drug delivery systems, vaccine carriers, biosensors, and coating materials. The polymerization of these lipid molecules forms a covalent bond between lipid moieties, which improves the noncovalent interactions that maintain the lipid lamellar phase architecture and increases the stability of the polymerized system. Because such lipid molecules form nanoassemblies with modifiable structures that acquire the stability of polymers following covalent bond formation, these lipids are of considerable interest in the emerging field of theranostics. In this Account, we summarize the biomedical applications of polymerizable lipids (primarily phospholipids) in the context of various nanoplatforms. We discuss stable nanoplatforms, which have been used in a variety of theranostics applications. In addition, we describe methods for assembling triggerable theranostics by combining appropriate nonpolymerizable lipids with polymerizable lipids. Polymeric lipids hold promise as nanotools in the field of medical imaging, targeting, and on-demand drug delivery. Because of their similarity to biological lipids, long-term toxicity issues from polymerizable lipid nanoplatforms are predicted to be minimal. Although the field of polymeric nanocapsules is still in development, intensive efforts are underway to produce systems which could be applied to disease diagnosis and treatment. We envision that nanoimaging platforms coupled with localized drug delivery technology will have a significant impact on cancer therapy and other related diseases. The existing wealth of clinical knowledge both in the photochemistry of imaging agents and/or drugs and modifications of these agents using light will prove valuable in the further development of polymeric theranostic lipid-based nanoparticles.     

Membranes for the life sciences and their future roles in medicine

Since the global outbreak of COVID-19, membrane technology for clinical treatments, including extracorporeal membrane oxygenation(ECMO) and protective masks and clothing, has attracted intense research attention for its irreplaceable abilities. Membrane research and applications are now playing an increasingly important role in various fields of life science. In addition to intrinsic properties such as size sieving,dissolution and diffusion, membranes are often endowed with additional functions ...     

Recent Applications of Deep Learning Methods on Evolution- and Contact-Based Protein Structure Prediction

The new advances in deep learning methods have influenced many aspects of scientific research, including the study of the protein system. The prediction of proteins’ 3D structural components is now heavily dependent on machine learning techniques that interpret how protein sequences and their homology govern the inter-residue contacts and structural organization. Especially, methods employing deep neural networks have had a significant impact on recent CASP13 and CASP14 competition. Here, we explore the recent applications of deep learning methods in the protein structure prediction area. We also look at the potential opportunities for deep learning methods to identify unknown protein structures and functions to be discovered and help guide drug–target interactions. Although significant problems still need to be addressed, we expect these techniques in the near future to play crucial roles in protein structural bioinformatics as well as in drug discovery.     

Zebrafish,Medaka and Turquoise Killifish for Understanding Human Neurodegenerative/Neurodevelopmental Disorders

In recent years, small fishes such as zebrafish and medaka have been widely recognized as model animals. They have high homology in genetics and tissue structure with humans and unique features that mammalian model animals do not have, such as transparency of embryos and larvae, a small body size and ease of experiments, including genetic manipulation. Zebrafish and medaka have been used extensively in the field of neurology, especially to unveil the mechanisms of neurodegenerative diseases such as Parkinson’s and Alzheimer’s disease, and recently, these fishes have also been utilized to understand neurodevelopmental disorders such as autism spectrum disorder. The turquoise killifish has emerged as a new and unique model animal, especially for ageing research due to its unique life cycle, and this fish also seems to be useful for age-related neurological diseases. These small fishes are excellent animal models for the analysis of human neurological disorders and are expected to play increasing roles in this field. Here, we introduce various applications of these model fishes to improve our understanding of human neurological disorders.     

Metagenomic Sequencing for Microbial DNA in Human Samples: Emerging Technological Advances

Whole genome metagenomic sequencing is a powerful platform enabling the simultaneous identification of all genes from entirely different kingdoms of organisms in a complex sample. This technology has revolutionised multiple areas from microbiome research to clinical diagnoses. However, one of the major challenges of a metagenomic study is the overwhelming non-microbial DNA present in most of the host-derived specimens, which can inundate the microbial signals and reduce the sensitivity of microorganism detection. Various host DNA depletion methods to facilitate metagenomic sequencing have been developed and have received considerable attention in this context. In this review, we present an overview of current host DNA depletion approaches along with explanations of their underlying principles, advantages and disadvantages. We also discuss their applications in laboratory microbiome research and clinical diagnoses and, finally, we envisage the direction of the further perfection of metagenomic sequencing in samples with overabundant host DNA.     

Piezoelectric glass-ceramics: Crystal chemistry,orientation mechanisms,and emerging applications

Piezoelectric materials have coupled mechanical and electrical energies and have long been used in devices for actuators, sensors, energy harvesters, frequency filters, and various additional applications. Piezoelectricity requires a non-centrosymmetric crystal structure and is therefore confined to materials that possess a periodic crystalline structure. Due to the non-crystalline nature of glass, piezoelectricity is fundamentally forbidden. However, one way to exploit piezoelectric properties in a glassy matrix is by developing glass-ceramics that possess controlled growth of a crystalline phase. Growth and orientation of piezoelectric crystals in a glassy matrix is a non-trivial process that has long been explored to combine the formability of glass with the thermal and mechanical resilience of glass-ceramics. While extensive work has been done in the field of functional glass-ceramics, the results are presented in isolated articles and a comprehensive review pertaining to symmetry breaking methods to exploit anisotropic properties in glass-ceramics has been absent from the literature. Here, we present a global review of the fundamental symmetry requirements for piezoelectricity, the development of polar, piezoelectric glass-ceramic compositions (specifically those with LiNbO₃ and fresnoite-based crystal phases), and various crystal growth and orientation mechanisms, including relevant kinetic and thermodynamic driving forces. Lastly, we discuss the challenges associated with implementing gradients to drive oriented crystal growth to develop non-centrosymmetry, and the need for future modeling work to produce adequate time-temperature-transformation (TTT) diagrams that take into account kinetic and thermodynamic driving forces for oriented crystal growth. Going beyond technical challenges, we conclude with an examination of current and potential applications for piezoelectric glass-ceramics that combine the formability of glass with the symmetry-dependent properties of ceramics.