Search

Methylation of benzene with methanol provides an efficient route to utilize the surplus benzene in aromatic united plants, facing a primary challenge from the ethylbenzene (EB) as the side product. This work discovers that EB selectivity can be significantly limited utilizing nanosized MFItype gallium silicate. Experiment and simulation results reveal diminished EB owing to the acid strength and diffusion capability. Gallium silicate zeolite, with much weaker acidity than aluminum silicate zeolite, favored higher energy barriers for polymethylbenzene formation, important precursors in methanol to olefin reactions, effectively suppressing ethene and EB. Additionally, the smaller particle size led to improved diffusion, facilitating the effusion of heavy aromatics. Therefore, nanosized MFItype gallium silicate could achieve an EB selectivity lower than 0.1% with toluene and xylene selectivities exceeding 90%. Our study provided a feasible strategy for adjusting the coordination–competition relationship between the main and side reaction mechanisms based on the synergistic effect of active sites and mass transfer.

With the continuous accumulation of massive amounts of mobile data, point-of-interest (POI) recommendation has become a vital task for location-based social networks. Deep neural networks or matrix factorization (MF) alone are challenging to effectively learn user–POI interaction functions. Moreover, the user–POI interaction matrix is sparse, and the heterogeneous characteristics of auxiliary information are underused. Therefore, we propose an innovative POI recommendation method that integrates attention-aware meta-paths based on deep neural matrix factorization (DNMF-AM). First, we develop a multi-relational heterogeneous information network of “user–POI–geographic region–POI category.” Multiple-weighted isomorphic information networks based on meta-paths are employed to obtain node-embedding vectors across different relationships. Attention networks are employed to aggregate node vectors across various relationships and serve as auxiliary information to mitigate the challenges of data sparsity. Subsequently, the internal embedding vectors of the users and POIs are extracted using feature embedding based on the user–POI interaction matrix. Second, these vectors are integrated with the embedding vectors obtained by aggregating the attention networks. Third, deep neural matrix factorization is used to learn linear and nonlinear user–POI interactions to mitigate the implicit feedback problem. This outcome is achieved using generalized matrix factorization and convolution-constrained multi-head self-attention mechanism deep neural networks. Extensive experiments conducted on two real-world datasets demonstrate that the DNMF-AM outperforms the optimal baseline NeuMF-CAA by 4.24% and 5.04% in terms of HR@10 and NDCG@10, respectively.

ObjectiveThe objective of this study is to describe the incidence of injuries and illnesses sustained during the Beijing Winter Olympic Games from 4 February 2022 to 20 February 2022.MethodsWe recorded the daily number of athlete injuries and illnesses (1) through the reporting of all National Olympic Committee (NOC) medical teams and (2) in the polyclinic and medical venues by the Beijing 2022 medical staff.ResultsIn total, 2848 athletes (1276 women, 45%; 1572 men, 55%) from 91 NOCs were followed prospectively for the occurrence of injury and illness. NOC and Beijing 2022 medical staff reported 289 injuries and 109 illnesses, equalling 10.1 injuries and 3.8 illnesses per 100 athletes over the 17-day period. The injury incidence was highest in ski halfpipe (30%), ski big air (28%), snowboard slopestyle (23%) and ski slopestyle (22%), and lowest (1%–2%) in curling, alpine mixed team parallel slalom, Nordic combined and alpine super-G. The illness incidence was highest in ski aerials (10%), skeleton (8%), cross-country skiing (8%) and Nordic combined (7%). In the study period, COVID-19 affected 32 athletes, accounting for 29% of all illnesses affecting 1.1% of all athletes.ConclusionOverall, 10% of the athletes incurred an injury and 4% an illness during the Beijing Winter Olympic Games. The incidence of illnesses overall, which was the lowest yet recorded in the Winter Olympic Games, and COVID-19 was mitigated through comprehensive countermeasures.

Transformer-based methods have demonstrated superior performance for monocular 3D object detection recently, which aims at predicting 3D attributes from a single 2D image. Most existing transformer-based methods leverage both visual and depth representations to explore valuable query points on objects, and the quality of the learned query points has a great impact on detection accuracy. Unfortunately, existing unsupervised attention mechanisms in transformers are prone to generate low-quality query features due to inaccurate receptive fields, especially on hard objects. To tackle this problem, this article proposes a novel “Supervised Scale-aware Deformable Attention” (SSDA) for monocular 3D object detection. Specifically, SSDA presets several masks with different scales and utilizes depth and visual features to adaptively learn a scale-aware filter for object query augmentation. Imposing the scale awareness, SSDA could well predict the accurate receptive field of an object query to support robust query feature generation. Aside from this, SSDA is assigned with a Weighted Scale Matching (WSM) loss to supervise scale prediction, which presents more confident results as compared to the unsupervised attention mechanisms. Extensive experiments on the KITTI and Waymo Open datasets demonstrate that SSDA significantly improves the detection accuracy, especially on moderate and hard objects, yielding state-of-the-art performance as compared to the existing approaches.

Traffic flow prediction is the key to transportation safety and efficiency. The advance in machine learning and deep learning has promoted the development of intelligent transportation systems. For example, the emergence of Deep Learning as a Service (DLaaS) has benefitted researchers a lot in dealing with large scale dataset and complex deep learning algorithms. In traffic forecasting, despite the success of deep learning-based models, there are still shortcomings, such as inadequate use of temporal and spatial traffic information, and indirect modeling of dependencies in traffic data. To address these challenges, we learn the transportation network in the form of a graph, and use graph wavelet as a key component to extract well-positioned features from the graph based on the transportation network. Compared with graph convolution, graph wavelets are very flexible and do not need to specify adjacent regions in the topological graph structure for feature extraction. At the same time, we propose to combine the multi-information fusion traffic control and guidance collaborative neural network and the results obtained are better than the benchmark algorithms. The results by comparison with several baseline methods show that our proposed method can outperform all the baseline methods.

High fluorescence background poses a substantial challenge to surface-enhanced Raman scattering (SERS), thereby limiting its broader applicability across diverse domains. In this work, silver nanoparticle (Ag NP)-loaded graphene oxide aerogel nanomaterials (GO-Ag ANM) were prepared for sensitive SERS detection of fluorescent explosive 2,4,8,10-tetranitrobenzo-1,3a,6,6a-tetraazapentaenopyridine (BPTAP) by a fluorescence quenching strategy. By harnessing the fluorescence quenching properties of graphene and the localized surface plasmon resonance of silver nanoparticles, the synthesized aerogels exhibited effective fluorescence quenching and Raman enhancement capabilities when employed for BPTAP analysis with 532 nm laser excitation. Significantly, precise control over the loading quantity of silver nanoparticles (Ag NPs) resulted in the remarkable sensitivity of the surface-enhanced Raman scattering (SERS) effect. This method allowed for the detection of fluorescent explosive BPTAP at an extraordinarily low concentration of 1 × 10−7M. Furthermore, the approach also demonstrated excellent detection capabilities for the dyes R6G, CV, and RhB. This study offers valuable insights for the sensitive detection of fluorescent molecules.

Using traffic data to accurately predict the traffic flow at a certain time in the future can alleviate problems such as traffic congestion, which plays an important role in the healthy transportation and economic development of cities. However, current traffic flow prediction models rely on human experience and only consider the advantages of single machine learning model. Therefore, in this work, we propose a multi-layer model based on transformer and deep learning for traffic flow prediction (MTDLTFP). The MTDLTFP model first draws on the idea of transformer model, which uses multiple encoders and decoders to perform feature extraction on the initial traffic data without human experience. In addition, in the prediction stage, the MTDLTFP model using deep learning technology, which input the hidden features into the convolutional neural network (CNN) and multi-layer feedforward neural network (MFNN) to obtain the prediction score respectively. The CNN model can captures the correlation information between the hidden features, and the MFNN can captures the nonlinear relationship between the features. Finally, we use a linear model to combine the two prediction scores, which can make the final prediction value take into account the common advantages of both models. Multiple experimental results on two real datasets demonstrate the effectiveness of the MTDLTFP model. The experimental results on the $WorkDay$ dataset are as follows, with the RMSE value of 0.191, MAE value of 0.165. The experimental results on the $HoliDay$ dataset are as follows, with RMSE value of 0.227, MAE value of 0.192.

Bearings are the core component of mechanical equipment. The health status of bearings is the key to the stable operation of the system. Bearing fault diagnosis model can discover damaged bearings in time, which has a large economic value for enterprises. The previous bearings fault diagnosis model suffers from problems such as small fault data and unrepresentative features, which leads to poor model generalization performance. Therefore, in this work, we propose a masked one-dimensional convolutional autoencoder (MOCAE) for bearing fault diagnosis based on digital twin enabled industrial internet of things (IIoT). The model monitors the bearing data using a set of IIoT platforms. The digital twin technology is used to build a digital twin model of the bearing device, and the parameters of the digital twin model are trained by the fault data obtained from the IIoT platform. The trained digital twin model can then simulate whether the bearing is faulty. In this digital twin model, MOCAE model is proposed for diagnosing faulty bearing signals. The MOCAE model first extracts the features from the time series signal of the bearing using a one-dimensional convolutional autoencoder, which can enhance the reconstruction ability of hidden features to make them more representative. Next, the MOCAE model automatically extracts the feature information contained in the time series signal data by self-training in order to reduce the dependence on the labeled data. The comprehensive experimental results on real bearing datasets show the superiority of the MOCAE model.

Chromatin-associated RNA (caRNA) acts as a ubiquitous epigenetic layer in eukaryotes, and has been reported to be essential in various biological processes. Here, we propose a highly interpretable computational framework, named DeepRCI, to identify the interactions between various types of RNAs and chromatin. DeepRCI can serve as a useful tool for characterizing RNA-chromatin interactions and studying the underlying gene regulatory code. Chromatin-associated RNA (caRNA) acts as a ubiquitous epigenetic layer in eukaryotes, and has been reported to be essential in various biological processes, including gene transcription, chromatin remodeling and cellular differentiation. Recently, numerous experimental techniques have been developed to characterize genome-wide RNA-chromatin interactions to understand their underlying biological functions. However, these experimental methods are generally expensive, time-consuming, and limited in identifying all potential sites, while most of the existing computational methods are restricted to detecting only specific types of RNAs interacting with chromatin. Here, we propose a highly interpretable computational framework, named DeepRCI, to identify the interactions between various types of RNAs and chromatin. In this framework, we introduce a novel deep learning component called variformer and integrate multi-omics data to capture intrinsic genomic features at both RNA and DNA levels. Extensive experiments demonstrate that DeepRCI can detect RNA-chromatin interactions more accurately when compared to the state-of-the-art baseline prediction methods. Furthermore, the sequence features extracted by DeepRCI can be well matched to known critical gene regulatory components, indicating that our model can provide useful biological insights into understanding the underlying mechanisms of RNA-chromatin interactions. In addition, based on the prediction results, we further delineate the relationships between RNA-chromatin interactions and cellular functions, including gene expression and the modulation of cell states. In summary, DeepRCI can serve as a useful tool for characterizing RNA-chromatin interactions and studying the underlying gene regulatory code.

In this paper, we consider the decoding of fountain codes where the received symbols may have errors. It is motivated by the application of fountain codes in DNA-based data storage systems where the inner code decoding, which generally has undetectable errors, is performed before the outer fountain code decoding. We propose a novel and efficient decoding algorithm, namely basis-finding algorithm (BFA), followed by three implementations. The key idea of the BFA is to find a basis of the received symbols, and then use the most reliable basis elements to recover the source symbols with the inactivation decoding. Gaussian elimination is used to find the basis and to identify the most reliable basis elements. As a result, the BFA has polynomial time complexity. For random fountain codes, we are able to derive some theoretical bounds for the frame error rate (FER) of the BFA. Extensive simulations with Luby transform (LT) codes show that, the BFA has significantly lower FER than the belief propagation (BP) algorithm except for an extremely large amount of received symbols, and the FER of the BFA generally decreases as the average weight of basis elements increases.

The flowshop sequence-dependent group scheduling problem (FSDGSP) with the production efficiency measures has been extensively studied due to its wide industrial applications. However, energy efficiency indicators are often ignored in the literature. This article considers the FSDGSP to minimize makespan, total flow time, and total energy consumption, simultaneously. After the problem-specific knowledge is extracted, a mixed-integer linear programming model and a critical path-based accelerated evaluation method are proposed. Since the FSDGSP includes multiple coupled subproblems, a greedy cooperative co-evolutionary algorithm (GCCEA) is designed to explore the solution space in depth. Meanwhile, a random mutation operator and a greedy energy-saving strategy are employed to adjust the processing speeds of machines to obtain a potential nondominated solution. A large number of experimental results show that the proposed algorithm significantly outperforms the existing classic multiobjective optimization algorithms, which is due to the usage of problem-related knowledge.

In this paper, we consider how to partition the parity-check matrices (PCMs) to reduce the hardware complexity and increase decoding throughput for the row layered decoding of quasi-cyclic low-density parity-check (QC-LDPC) codes. First, we formulate the PCM partitioning as an optimization problem, which targets to minimize the maximum column weight of each layer while maintaining a block cyclic shift property among different layers. As a result, we derive all the feasible solutions for the problem and propose a tight lower bound $\omega _{LB}$ on the minimum possible maximum column weight to evaluate a solution. Second, we define a metric called layer distance to measure the data dependency between consecutive layers and further illustrate how to identify the solutions with desired layer distance from those achieving the minimum value of $\omega _{LB}=1$ , which is preferred to reduce computation delay. Next, we demonstrate that up-to-now, finding an optimal solution for the optimization problem with polynomial time complexity is unachievable. Therefore, both enumerative and greedy partition algorithms are proposed instead. After that, we modify the quasi-cyclic progressive edge-growth (QC-PEG) algorithm to directly construct PCMs that have a straightforward partition scheme to achieve $\omega _{LB} $ or the desired layer distance. Simulation results showed that the constructed codes have better error correction performance and achieve less average number of iterations than the underlying 5G LDPC codes.

Bone grafting is frequently conducted to treat bone defects caused by trauma and tumor removal, yet with significant medical and socioeconomic burdens. Space-occupying bone substitutes remain challenging in the control of osteointegration, and meanwhile activation of endogenous periosteal cells by using non-space-occupying implants to promote new bone formation becomes another therapeutic strategy. Here, we fabricated a magnesium-based artificial bandage with optimal micropatterns for activating periosteum-associated biomineralization. Collagen was self-assembled on the surface of magnesium oxide nanoparticles embedded electrospun fibrous membranes as a hierarchical bandage structure to facilitate the integration with periosteum in situ. After the implantation on the surface of cortical bone in vivo, magnesium ions were released to generate a pro-osteogenic immune microenvironment by activating the endogenous periosteal macrophages into M2 phenotype and, meanwhile, promote blood vessel formation and neurite outgrowth. In a cortical bone defect model, magnesium-based artificial bandage guided the surrounding newly formed bone tissue to cover the defected area. Taken together, our study suggests that the strategy of stimulating bone formation can be achieved with magnesium delivery to periosteum in situand the proposed periosteal bandages act as a bioactive media for accelerating bone healing.

Hypoxia at the tumor site limits the therapeutic effects of photodynamic therapy (PDT) in oral squamous cell carcinoma (OSCC), which is an oxygen-consumption process. Inhibiting cellular oxygen consumption and reducing cellular ATP production are expected to enhance PDT. In this study, we designed and constructed dandelion-like size-shrinkable nanoparticles for tumor-targeted delivery of hypoxia regulator resveratrol (RES) and photodynamic agent chlorine e6 (CE6). Both drugs were co-encapsulated in small-sized micelles modified with EGFR targeting ligand GE11, which was further conjugated on hyaluronic nanogel (NG) to afford RC-GMN. After targeted accumulation in tumors mediated by GE11 and enhanced penetration and retention (EPR) effects, RC-GMN was degraded by hyaluronidase (HAase) and resulted in small-sized micelles, allowing for deep penetration and dual-receptor-mediated cellular internalization. Resveratrol inhibited cellular oxygen consumption and provided sufficient oxygen for PDT, which consequently activated PDT to produce reactive oxygen species (ROS). Notably, we found that autophagy was overactivated in PDT, which was further strengthened by the hypoxia regulator resveratrol, elevating autophagic cell death. The synergistic effects of resveratrol and CE6 promoted autophagic cell death and apoptosis in the enhanced PDT, resulting in stronger antitumor effects in the orthotopic OSCC model. Therefore, the facilitated delivery of hypoxia regulator enhanced PDT efficacy by elevating oxygen content in tumor cells and inducing autophagic cell death and apoptosis, which offers an alternative strategy for enhancing the PDT effects against OSCC.

Guided bone regeneration (GBR) technique using a barrier membrane holds great potential to allow the single-stage reconstruction of critical-sized bone defects. Here, bioactive nanoneedle-like magnesium oxychloride ceramics (MOCs) are synthesized and recruited as an osteoinductive factor within a polycaprolactone-gelatin A (PCL-GelA) membranous matrix to generate a periosteum-mimicking biphasic GBR membrane (PCL-GelA/MOC) to accelerate calvarial defect repair. The PCL-GelA/MOC membrane acts as a shield for defect areas and a reservoir of osteoinductive molecules, which provides a favorable microenvironment for supporting cell proliferation, infiltration, and differentiation. This membrane leads to accelerated osteogenesis and angiogenesis, effectual defect bridging, and significantly enhanced bone regeneration when applied to a 5 mm sized rat calvarial defect. This makes this innovative and multifunctional GBR membrane a suitable candidate for clinical applications with promising curative efficacy.

The embryonic ectoderm development (EED) is a core component of the polycomb-repressive complex 2 (PRC2) whose mutations are linked to neurodevelopmental abnormalities, intellectual disability, and neurodegeneration. Although EED has been extensively studied in neural stem cells and oligodendrocytes, its role in microglia is incompletely understood. Here, we show that microglial EED is essential for synaptic pruning during the postnatal stage of brain development. The absence of microglial EED at early postnatal stages resulted in reduced spines and impaired synapse density in the hippocampus at adulthood, accompanied by upregulated expression of phagocytosis-related genes in microglia. As a result, deletion of microglial Eedimpaired hippocampus-dependent learning and memory in mice. These results suggest that microglial EED is critical for normal synaptic and cognitive functions during postnatal development.

Chronic diabetic wound healing remains a challenge due to the existence of excessive danger molecules and bacteria in the inflammatory microenvironment. There is an urgent need for advanced wound dressings that target both inflammation and infection. Here, a bioactive hydrogel without loading any anti-inflammatory ingredients is rationally designed to achieve a “Pull–Push” approach for efficient and safe bacteria-infected diabetic wound healing by integrating danger molecule scavenging (Pull) with antibiotic delivery (Push) in the inflammatory microenvironment. The cationic hydrogel, termed the OCMC-Tob/PEI hydrogel, is fabricated by the conjugation of polyethylenimine (PEI) and tobramycin (Tob) on an oxidized carboxymethyl cellulose (OCMC) backbone via the Schiff base reaction with injectable, self-healing, and biocompatible properties. The OCMC-Tob/PEI hydrogel not only displays the remarkable capability of capturing multiple negatively charged danger molecules (e.g., cell-free DNA, lipopolysaccharides, and tumor necrosis factor-α) to ameliorate anti-inflammation effects but also achieves controllable long-term antibacterial activity by the pH-sensitive release of Tob. Consequently, this multifunctional hydrogel greatly expedites the wound closure rate with combined anti-inflammation and anti-infection effects on Pseudomonas aeruginosa-infected diabetic wounds. Our work provides a highly versatile treatment approach for chronic diabetic wounds and a promising dressing for regenerative medicine.

The GDSL esterase/lipase family contains many functional genes that perform important biological functions in growth and development, morphogenesis, seed oil synthesis, and defense responses in plants. The expression of GDSL esterase/lipase genes can respond to biotic and abiotic stresses. Although GDSL esterase/lipase family genes have been identified and studied in other plants, they have not been identified and their functions remain unclear in tomato. This study is the first to identify 80 GDSL esterase/lipase family genes in tomato, which were named SlGELP1–80. These genes were mapped to their positions on the chromosomes and their physical and chemical properties, gene structure, phylogenetic relationships, collinear relationships, and cis-acting elements were analyzed. The spatiotemporal expression characteristics of the SlGELPgenes in tomato were diverse. In addition, RNA-seq analysis indicated that the expression patterns of the SlGELPgenes in tomato differed before and after inoculation with Stemphylium lycopersici. qRT-PCR was used to analyze the expression of five SlGELPgenes after treatments with S. lycopersici, salicylic acid and jasmonic acid. Finally, this study was the first to identify and analyze GDSL esterase/lipase family genes in tomato viabioinformatics approaches, and these findings provide new insights for improving the study of plant disease resistance.

Background:: Mandibular distraction osteogenesis (MDO) is a kind of endogenous tissue engineering technology that lengthens the jaw and opens airway so that a patient can breathe safely and comfortably on his or her own. Endothelial progenitor cells (EPCs) are crucial for MDO-related angiogenesis. Moreover, emerging evidence suggests that heat shock protein 20 (Hsp20) modulates angiogenesis under hypoxic conditions. However, the specific role of Hsp20 in EPCs, in the context of MDO, is not yet known. The aim of this study was to explore the expression of Hsp20 during MDO and the effects of Hsp20 on EPCs under hypoxia. Methods:: Mandibular distraction osteogenesis and mandibular bone defect (MBD) canine model were established. The expression of CD34, CD133, HIF-1α, and Hsp20 in callus was detected by immunofluorescence on day 14 after surgery. Canine bone marrow EPCs were cultured, with or without optimal cobalt chloride (CoCl2) concentration. Hypoxic effects, caused by CoCl2, were evaluated by means of the cell cycle, cell apoptosis, transwell cell migration, and tube formation assays. The Hsp20/KDR/PI3K/Akt expression levels were evaluated via immunofluorescence, RT-qPCR, and western blot. Next, EPCs were incorporated with either Hsp20-overexpression or Hsp20-siRNA lentivirus. The resulting effects were evaluated as described above. Results:: CD34, CD133, HIF-1α, and Hsp20 were displayed more positive in the callus of MDO compared with MBD. In addition, hypoxic conditions, generated by 0.1 mM CoCl2, in canine EPCs, accelerated cell proliferation, migration, tube formation, and Hsp20 expression. Hsp20 overexpression in EPCs significantly stimulated cell proliferation, migration, and tube formation, whereas Hsp20 inhibition produced the opposite effect. Additionally, the molecular mechanism was partly dependent on the KDR/PI3K/Akt pathway. Conclusion:: In summary, herein, we present a novel mechanism of Hsp20-mediated regulation of canine EPCs via Akt activation in a hypoxic microenvironment.

Compared to other commercial atomic clocks in the time keeping field, the greatest advantage of cesium beam atomic clocks is their superior long-term stability. Compared to magnetic state-selection clocks, optically pumped cesium beam atomic clocks have more interacting atoms, which results in better stability potential. To achieve good long-term stability, we propose methods including stabilization of laser power and reconstruction of circuits. They play a key role in the long-term stability of cesium beam atomic clocks. After 75 days of continuous running and measurement, we released the 5-day stability results (7×10^−15 Allan deviation) of our optically pumped cesium beam atomic clock. To the best of our knowledge, this is the best 5-day stability result ever reported for compact optically pumped cesium beam atomic clocks.

Abstract: A novel tetraphenylethylene (TPE)-functionalized arylimidazole derivative (HTPEAIM) has been designed and synthesized. The structure was characterized by FT-IR, HR-MS, 1H and 13C NMR. The photophysical properties were investigated by UV–vis absorption and PL spectra. HTPEAIMexhibited excellent aggregation-induced emission (AIE) property that it was non-emission in organic solution and with high fluorescence in aggregation state as the restriction of intramolecular motions (RIMs). Moreover, HTPEAIMcould be a fluorogenic probe for sensing picric acid (PA) and Cu2+in the aqueous environment. Since the proton binding capacity of the N atom of the imidazole group and the donor–acceptor interactions, HTPEAIMwas able to sensitively and selectively detect PA through the synergistic effects of photoinduced energy transfer (PET) and fluorescence resonance energy transfer (FRET) with a limit of detection (LOD) of 0.99 μM. Meanwhile, driving by the metal binding ability of the pincer-like structure of hydroxyl and imidazole group, HTPEAIMwas used for sensing Cu2+in the aqueous environment. The response mechanism was systemically studied, and the results showed that the stable Cu(HTPEAIM)2complex was formed and that induced the fluorescence highly quenching to obtain a low LOD of 34.8 nM. Graphical Abstract:

The application of artificial intelligence (AI) technology in the medical field has experienced a long history of development. In turn, some long-standing points and challenges in the medical field have also prompted diverse research teams to continue to explore AI in depth. With the development of advanced technologies such as the Internet of Things (IoT), cloud computing, big data, and 5G mobile networks, AI technology has been more widely adopted in the medical field. In addition, the in-depth integration of AI and IoT technology enables the gradual improvement of medical diagnosis and treatment capabilities so as to provide services to the public in a more effective way. In this work, we examine the technical basis of IoT, cloud computing, big data analysis and machine learning involved in clinical medicine, combined with concepts of specific algorithms such as activity recognition, behavior recognition, anomaly detection, assistant decision-making system, to describe the scenario-based applications of remote diagnosis and treatment collaboration, neonatal intensive care unit, cardiology intensive care unit, emergency first aid, venous thromboembolism, monitoring nursing, image-assisted diagnosis, etc. We also systematically summarize the application of AI and IoT in clinical medicine, analyze the main challenges thereof, and comment on the trends and future developments in this field.

Pseudouridine (Ψ) is the most prevalent post-transcriptional RNA modification and is widespread in small cellular RNAs and mRNAs. However, the functions, mechanisms, and precise distribution of Ψs (especially in mRNAs) still remain largely unclear. The landscape of Ψs across the transcriptome has not yet been fully delineated. Here, we present a highly effective model based on a convolutional neural network (CNN), called PseudoUridyLation Site Estimator (PULSE), to analyze large-scale profiling data of Ψ sites and characterize the contextual sequence features of pseudouridylation. PULSE, consisting of two alternatively-stacked convolution and pooling layers followed by a fully-connected neural network, can automatically learn the hidden patterns of pseudouridylation from the local sequence information. Extensive validation tests demonstrated that PULSE can outperform other state-of-the-art prediction methods and achieve high prediction accuracy, thus enabling us to further characterize the transcriptome-wide landscape of Ψ sites. We further showed that the prediction results derived from PULSE can provide novel insights into understanding the functional roles of pseudouridylation, such as the regulations of RNA secondary structure, codon usage, translation, and RNA stability, and the connection to single nucleotide variants. The source code and final model for PULSE are available at https://github.com/mlcb-thu/PULSE.

Huntington’s disease (HD) is an autosomal dominant neurodegenerative disorder caused by toxic aggregates of mutant huntingtin protein (mHTT) in the brain. Decreasing mHTT is a potential strategy for therapeutic purpose of HD. Valosin-containing protein (VCP/p97) is a crucial regulator of proteostasis, which regulates the degradation of damaged protein through proteasome and autophagy pathway. Since VCP has been implicated in pathogenesis of HD as well as other neurodegenerative diseases, small molecules that specifically regulate the activity of VCP may be of therapeutic benefits for HD patients. In this study we established a high-throughput screening biochemical assay for VCP ATPase activity measurement and identified gossypol, a clinical approved drug in China, as a novel modulator of VCP. Gossypol acetate dose-dependently inhibited the enzymatic activity of VCP in vitro with IC50of 6.53±0.6 μM. We further demonstrated that gossypol directly bound to the interface between the N and D1 domains of VCP. Gossypol acetate treatment not only lowered mHTT levels and rescued HD-relevant phenotypes in HD patient iPS-derived Q47 striatal neurons and HD knock-in mouse striatal cells, but also improved motor function deficits in both Drosophilaand mouse HD models. Taken together, gossypol acetate acted through a gain-of-function way to induce the formation of VCP-LC3-mHTT ternary complex, triggering autophagic degradation of mHTT. This study reveals a new strategy for treatment of HD and raises the possibility that an existing drug can be repurposed as a new treatment of neurodegenerative diseases.

Pseudouridine (Ψ) is the most prevalent post-transcriptional RNA modification and is widespread in small cellular RNAs and mRNAs. However, the functions, mechanisms, and precise distribution of Ψs (especially in mRNAs) still remain largely unclear. The landscape of Ψs across the transcriptome has not yet been fully delineated. Here, we present a highly effective model based on a convolutional neural network (CNN), called PseudoUridyLation Site Estimator (PULSE), to analyze large-scale profiling data of Ψ sites and characterize the contextual sequence features of pseudouridylation. PULSE, consisting of two alternatively-stacked convolution and pooling layers followed by a fully-connected neural network, can automatically learn the hidden patterns of pseudouridylation from the local sequence information. Extensive validation tests demonstrated that PULSE can outperform other state-of-the-art prediction methods and achieve high prediction accuracy, thus enabling us to further characterize the transcriptome-wide landscape of Ψ sites. We further showed that the prediction results derived from PULSE can provide novel insights into understanding the functional roles of pseudouridylation, such as the regulations of RNA secondary structure, codon usage, translation, and RNA stability, and the connection to single nucleotide variants. The source code and final model for PULSE are available at https://github.com/mlcb-thu/PULSE.

The adaptive sparse representation (ASR) can effectively combine the structure information of a sample dictionary and the sparsity of coding coefficients. This algorithm can effectively consider the correlation between training samples and convert between sparse representation-based classifier (SRC) and collaborative representation classification (CRC) under different training samples. Unlike SRC and CRC which use fixed norm constraints, ASR can adaptively adjust the constraints based on the correlation between different training samples, seeking a balance between l1 and l2 norm, greatly strengthening the robustness and adaptability of the classification algorithm. The correlation coefficients (CC) can better identify the pixels with strong correlation. Therefore, this article proposes a hyperspectral image classification method called correlation coefficients and adaptive sparse representation (CCASR), based on ASR and CC. This method is divided into three steps. In the first step, we determine the pixel to be measured and calculate the CC value between the pixel to be tested and various training samples. Then we represent the pixel using ASR and calculate the reconstruction error corresponding to each category. Finally, the target pixels are classified according to the reconstruction error and the CC value. In this article, a new hyperspectral image classification method is proposed by fusing CC and ASR. The method in this paper is verified through two sets of experimental data. In the hyperspectral image (Indian Pines), the overall accuracy of CCASR has reached 0.9596. In the hyperspectral images taken by HIS-300, the classification results show that the classification accuracy of the proposed method achieves 0.9354, which is better than other commonly used methods.