Search

Studying the flow behavior is critical to understand the deformation mechanism of amorphous solids. However, detecting the basic flow events in amorphous solids is challenging. Here, by simultaneous SAXS/WAXS, elementary flow carriers in wound metallic glasses are identified from flow-induced structural heterogeneities with a radius of gyration of 2.5∼3.5 nm. Their size increases and morphology changes from sphere-like to rod-like under flow. Moreover, the atomic structure exhibits an unusual change to a more disordered state during winding/annealing at the temperature of ∼0.8 Tg. This work provides an atomic-to-nanoscale description of the flow carriers of amorphous solids during deformation. Impact Statement The elementary flow carriers in metallic glasses have a radius of gyration of 2.5∼3.5 nm. During flow, their size increases and morphology changes from sphere-like to rod-like.

Early prediction of Alzheimer's disease (AD) is crucial for timely intervention and treatment. This study aims to use machine learning approaches to analyze longitudinal electronic health records (EHRs) of patients with AD and identify signs and symptoms that can predict AD onset earlier. We used a case-control design with longitudinal EHRs from the U.S. Department of Veterans Affairs Veterans Health Administration (VHA) from 2004 to 2021. Cases were VHA patients with AD diagnosed after 1/1/2016 based on ICD-10-CM codes, matched 1:9 with controls by age, sex and clinical utilization with replacement. We used a panel of AD-related keywords and their occurrences over time in a patient's longitudinal EHRs as predictors for AD prediction with four machine learning models. We performed subgroup analyses by age, sex, and race/ethnicity, and validated the model in a hold-out and "unseen" VHA stations group. Model discrimination, calibration, and other relevant metrics were reported for predictions up to ten years before ICD-based diagnosis. The study population included 16,701 cases and 39,097 matched controls. The average number of AD-related keywords (e.g., "concentration", "speaking") per year increased rapidly for cases as diagnosis approached, from around 10 to over 40, while remaining flat at 10 for controls. The best model achieved high discriminative accuracy (ROCAUC 0.997) for predictions using data from at least ten years before ICD-based diagnoses. The model was well-calibrated (Hosmer-Lemeshow goodness-of-fit p-value = 0.99) and consistent across subgroups of age, sex and race/ethnicity, except for patients younger than 65 (ROCAUC 0.746). Machine learning models using AD-related keywords identified from EHR notes can predict future AD diagnoses, suggesting its potential use for identifying AD risk using EHR notes, offering an affordable way for early screening on large population., Comment: 24 pages

In the Himalayas, extreme precipitation events are rare but recurrent, leading to severe landslides and flooding. Although these remote mountain regions are sparsely populated, natural disasters are often accompanied by significant socio-economic consequences. In recent years, for example in 2014 and 2022, extreme precipitation events have been observed, even in the autumn months when precipitation is usually low. The high precipitation amounts and cool temperatures lead to an unusually high snow cover at high altitudes. In this study, we assess the impact of these snowpacks on the climatic mass balance of mountain glaciers in the Nepalese Himalayas. On the one hand, the insulation effect can be important, but also the increased albedo, which influences the development of the mass balance. We force a glacier energy and mass balance model with reanalysis data using an automatic weather station at Halji Glacier in northwestern Nepal for bias correction. The impacts are quantified by simulating the annual mass balance with the original forcing dataset and with an artificial dataset excluding unusual precipitation events. Furthermore, it is known that snowfall in May and June, at the beginning of the monsoon season, has a great influence on the ablation season, since in addition to the mass input, the albedo also increases, which leads to less absorbed short-wave radiation. Therefore, in addition to the extreme events in autumn, the influence of extreme events during this period will also be investigated., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

Opioid related aberrant behaviors (ORAB) present novel risk factors for opioid overdose. Previously, ORAB have been mainly assessed by survey results and by monitoring drug administrations. Such methods however, cannot scale up and do not cover the entire spectrum of aberrant behaviors. On the other hand, ORAB are widely documented in electronic health record notes. This paper introduces a novel biomedical natural language processing benchmark dataset named ODD, for ORAB Detection Dataset. ODD is an expert-annotated dataset comprising of more than 750 publicly available EHR notes. ODD has been designed to identify ORAB from patients' EHR notes and classify them into nine categories; 1) Confirmed Aberrant Behavior, 2) Suggested Aberrant Behavior, 3) Opioids, 4) Indication, 5) Diagnosed opioid dependency, 6) Benzodiapines, 7) Medication Changes, 8) Central Nervous System-related, and 9) Social Determinants of Health. We explored two state-of-the-art natural language processing (NLP) models (finetuning pretrained language models and prompt-tuning approaches) to identify ORAB. Experimental results show that the prompt-tuning models outperformed the finetuning models in most cateogories and the gains were especially higher among uncommon categories (Suggested aberrant behavior, Diagnosed opioid dependency and Medication change). Although the best model achieved the highest 83.92\% on area under precision recall curve, uncommon classes (Suggested Aberrant Behavior, Diagnosed Opioid Dependence, and Medication Change) still have a large room for performance improvement., Under review

The characteristics of the study population were described in Supplementary Table 1. Details of SNPs for urate, hand grip strength and ALM are shown in Supplementary Tables 2-5. Scatter plots, funnel plots and leave-one-out sensitivity analysis are presented in Supplementary Fig. 1-3.

Considering the pile-slab subgrade project of the Hangzhou-Huang Shan Passenger Dedicated Line as the basis, this paper conducts a 1:10 large-scale indoor model test for the horizontal bearing capacity of the pile-slab structure in steep mountainous areas to study the distribution of the pile-slab structure stress, soil pressure and structural deformation and analyze the failure mode of the structure and slope. The research shows that when the subgrade with a double-row pile-slab structure is subjected to horizontal loading in the steep slope section, the steel bars of the pile body above the sliding surface are compressed, and the steel bars of the pile body below the sliding surface are under tension. With the increase in the horizontal load, the stress of the pile body steel bar remains basically unchanged or shows a steady increase and finally sharply increases. The deformation of the bearing plate isdominated by the horizontal displacement, and the horizontal displacement reaches 7.25 mm when the plate is broken. In addition, warping deformation of the inner high and outer low occurs. When the horizontal load reaches 157 kN, shallow damage and local collapse of the slope occur, and transverse and diagonal cracks occur at the top of the pile and near the sliding surface of the pile. During the test, the pile-slab structure always deforms more than the slope, and the overall stability of the structure is good. The test is suitable for sections where the remaining sliding force is less than 770 kN/m (equivalent to a slope length of 79.123 m).

Summarizing lengthy documents is a common and essential task in our daily lives. Although recent advancements in neural summarization models can assist in crafting general-purpose summaries, human writers often have specific requirements that call for a more customized approach. To address this need, we introduce REVISE (Refinement and Editing via Iterative Summarization Enhancement), an innovative framework designed to facilitate iterative editing and refinement of draft summaries by human writers. Within our framework, writers can effortlessly modify unsatisfactory segments at any location or length and provide optional starting phrases -- our system will generate coherent alternatives that seamlessly integrate with the existing summary. At its core, REVISE incorporates a modified fill-in-the-middle model with the encoder-decoder architecture while developing novel evaluation metrics tailored for the summarization task. In essence, our framework empowers users to create high-quality, personalized summaries by effectively harnessing both human expertise and AI capabilities, ultimately transforming the summarization process into a truly collaborative and adaptive experience.

IntroductionThe oral cavity and the gut tract are interconnected, and both contain abundant natural microbiota. Gut microbiota may interact with oral flora and participate in the development of periodontitis. However, the specific role of certain gut microbiota taxa for periodontitis has not been investigated. Mendelian Randomization is an ideal method to explore causal relationships avoiding reverse causality and potential confounding factors. Thus, we conducted a two-sample Mendelian Randomization study to comprehensively reveal the potential genetic causal effect of gut microbiota on periodontitis.MethodsSNPs strongly associated with 196 gut microbiota taxa (18,340 individuals) were selected as instrument variables, and periodontitis (17,353 periodontitis cases and 28,210 controls) was used as the outcome. The causal effect was analyzed via random effect inverse variance-weighted, weighted median, and MR-Egger. The sensitivity analyses were conducted using Cochran’s Q tests, funnel plots, leave-one-out analyses, and MR-Egger intercept tests.ResultsNine gut microbiota taxa (Prevotella 7, Lachnospiraceae UCG-008, Enterobacteriales, Pasteurellales, Enterobacteriaceae, Pasteurellaceae, Bacteroidales S24.7 group, Alistipes, and Eisenbergiella) are predicted to play a causal role in enhancing the risk of periodontitis (p< 0.05). Besides, two gut microbiota taxa (Butyricicoccus and Ruminiclostridium 6) have potentially inhibitive causal effects on the risk of periodontitis (p< 0.05). No significant estimation of heterogeneity or pleiotropy is detected.ConclusionOur study demonstrates the genetic causal effect of 196 gut microbiota taxa on periodontitis and provides guidance for the clinical intervention of periodontitis.

Sweet orange ‘Newhall’ peels (SOPs) are abundant in flavonoids, making them increasingly popular in the realms of nutrition, food, and medicine. However, there is still much unknown about flavonoid components in SOPs and the molecular mechanism of flavonoid biosynthesis when subjected to magnesium stress. The previous experiment conducted by the research group found that the total flavonoid content of Magnesium deficiency (MD) was higher than Magnesium sufficiency (MS) in SOPs. In order to study the metabolic pathway of flavonoids under magnesium stress, an integrative analysis of the metabolome and transcriptome was performed in SOPs at different developmental stages, comparing MS and MD. A comprehensive analysis revealed the identification of 1,533 secondary metabolites in SOPs. Among them, 740 flavonoids were classified into eight categories, with flavones identified as the dominant flavonoid component. The influence of magnesium stress on flavonoid composition was evaluated using a combination of heat map and volcanic map, which indicated significant variations between MS and MD varieties at different growth stages. The transcriptome detected 17,897 differential genes that were significantly enriched in flavonoid pathways. Further analysis was performed using Weighted gene correlation network analysis (WGCNA) in conjunction with flavonoid metabolism profiling and transcriptome analysis to identify six hub structural genes and ten hub transcription factor genes that play a crucial role in regulating flavonoid biosynthesis from yellow and blue modules. The correlation heatmap and Canonical Correspondence Analysis (CCA) results showed that CitCHS had a significant impact on the synthesis of flavones and other flavonoids in SOPs, as it was the backbone gene in the flavonoid biosynthesis pathway. The qPCR results further validated the accuracy of transcriptome data and the reliability of candidate genes. Overall, these results shed light on the composition of flavonoid compounds in SOPs and highlight the changes in flavonoid metabolism that occur under magnesium stress. This research provides valuable insights for improving the cultivation of high-flavonoid plants and enhancing our understanding of the molecular mechanisms underlying flavonoid biosynthesis.

We present an empirical evaluation of various outputs generated by nine of the most widely-available large language models (LLMs). Our analysis is done with off-the-shelf, readily-available tools. We find a correlation between percentage of memorized text, percentage of unique text, and overall output quality, when measured with respect to output pathologies such as counterfactual and logically-flawed statements, and general failures like not staying on topic. Overall, 80.0% of the outputs evaluated contained memorized data, but outputs containing the most memorized content were also more likely to be considered of high quality. We discuss and evaluate mitigation strategies, showing that, in the models evaluated, the rate of memorized text being output is reduced. We conclude with a discussion on potential implications around what it means to learn, to memorize, and to evaluate quality text., Preprint. Under review

Throughout its history, Operational Research has evolved to include a variety of methods, models and algorithms that have been applied to a diverse and wide range of contexts. This encyclopedic article consists of two main sections: methods and applications. The first aims to summarise the up-to-date knowledge and provide an overview of the state-of-the-art methods and key developments in the various subdomains of the field. The second offers a wide-ranging list of areas where Operational Research has been applied. The article is meant to be read in a nonlinear fashion. It should be used as a point of reference or first-port-of-call for a diverse pool of readers: academics, researchers, students, and practitioners. The entries within the methods and applications sections are presented in alphabetical order. The authors dedicate this paper to the 2023 Turkey/Syria earthquake victims. We sincerely hope that advances in OR will play a role towards minimising the pain and suffering caused by this and future catastrophes.

Recent years have witnessed an increase in applications of artificial intelligence (AI) in the detection of oceanic features with the tremendous success of deep learning. Given the unique biological ecosystems and mineral-rich deposits, the exploration of hydrothermal fields is both scientifically and commercially important. To achieve autonomous and intelligent sampling of the hydrothermal plume by using AUV, this paper proposes an innovative method for real-time plume detection based on the YOLOv5n deep learning algorithm designed with a light-weight neural network architecture to meet the requirements of embedded platforms. Ground truth labeler app LabelImg was used to generate the ground truth data from the plume dataset created by ourselves. To accurately and efficiently detect hydrothermal plumes using an embedded system, we improved the original structure of YOLOv5n in two aspects. First, SiLU activation functions in the model were replaced by ReLU activations at shallow layers and Hard-SiLU activations at deep layers to reduce the number of calculations. Second, an attention module termed Coordinate Attention (CA) was integrated into the model to improve its sensitivity to both channel and spatial features. In addition, a transfer learning training method was adopted to further improve the model’s accuracy and generalizability. Finally, we successfully deployed the proposed model in a low-cost embedded device (NVIDIA Jetson TX2 NX) by using the TensorRT inference engine. We then installed the Jetson TX2 NX into a hovering-type AUV as its vision processing unit and conducted a plume detection test in the water tank. The water tank experimental results demonstrated that the proposed method can achieve real-time onboard hydrothermal plume detection.

Additional file 1: Supplementary Figure 1. Multiple sequence alignment and amino acid conservation of MdCYP proteins. Supplementary Figure 2. All protein domains of MdCYPs. Supplementary Figure 3. Amino acid sequences and conservation of 15 motifs. Supplementary Figure 4. Secondary structure statistics of MdCYP proteins. Supplementary Figure 5. Collinearity analysis of MdCYPs in the apple genome-wide context. Supplementary Figure 6. Protein domains of 5 CYPs in CladeIII. Supplementary Table 1. Information of CYPs in Malus (M. sieversii, M. sylvestris, M. baccata and M. prunifolia). Supplementary Table 2. Information statistics of MdCYP family protein homology modeling. Supplementary Table 3. Segmental duplication genes in colinear gene pairs of MdCYPs. Supplementary Table 4. The intergenomic duplications between. Supplementary Table 5. The qRT-PCR primer sequences of MdCYPs. Supplementary Table 6. Primer of coding sequence and promoter sequence of MdCYP16.

The Tibetan Plateau (TP) is called the “Third Pole”, which contains the largest amount of snow and glaciers outside the polar regions. As an important and variable feature of the land surface, snow coverage on the TP has great impacts on regional climates. However, the commonly used ERA5 reanalysis in dynamic downscaling largely overestimates the snow depth for the TP. To improve the representation of snow cover in ERA5, a new ERA5 driven downscaling dataset (High Asia Refined analysis version 2, HAR v2) was generated by the Weather Research and Forecasting (WRF) model with the bias-corrected snow depth. This study aims to identify and better understand the impacts of bias-corrected initial snow conditions on dynamical downscaling simulations, by comparing the HAR v2 with a 5-year ERA5 forced WRF simulation without bias correction of initial snow depth (referred to as WRF_ERA5). Results show that the bias correction significantly improves the simulation of air temperature at 2 m (T2), with regional mean cold bias reduced by 0.2-2.4°C, but no significant improvement in precipitation simulation is found. Further comparative analysis reveals that higher snow depth in WRF_ERA5 leads to T2, mean daily precipitation, summer extreme precipitation, and contributions of convective precipitation to summer mean daily precipitation decrease by 0-4°C, 0-60%, 0-40%, and 0-10%, respectively, in most areas of the TP. Overall, this study confirms the importance of snow cover for the TP climate., The 28th IUGG General Assembly (IUGG2023) (Berlin 2023)

In this paper we establish an asymptotic formula for the number of rational points of bounded anticanonical height which lie on a certain Zariski dense subset of the biprojective hypersurface \begin{align*} x_1y_1^2+...+x_sy_s^2 = 0 \end{align*} in $\mathbb{P}^{s-1} \times \mathbb{P}^{s-1} $ with $s \geq 7$. This confirms the Manin conjecture for this variety., Comment: Added Appendix to address norm issues in Lemma 2.2

Fe-based metallic glasses are promising functional materials for advanced magnetism and sensor fields. Tailoring magnetic performance in amorphous materials requires a thorough knowledge of the correlation between structural disorder and magnetic order, which remains ambiguous. Two practical difficulties remain: the first is directly observing subtle magnetic structural changes on multiple scales, and the second is precisely regulating the various amorphous states. Here we propose a novel approach to tailor the amorphous structure through the liquid liquid phase transition. In-situ synchrotron diffraction has unraveled a medium-range ordering process dominated by edge-sharing cluster connectivity during the liquid-liquid phase transition. Moreover, nanodomains with topological order have been found to exist in composition with liquid-liquid phase transition, manifesting as hexagonal patterns in small-angle neutron scattering profiles. The liquid-liquid phase transition can induce the nanodomains to be more locally ordered, generating stronger exchange interactions due to the reduced Fe-Fe bond and the enhanced structural order, leading to the increment of saturation magnetization. Furthermore, the increased local heterogeneity in the medium range scale enhances the magnetic anisotropy, promoting the permeability response under applied stress and leading to a better stress-impedance effect. These experimental results pave the way to tailor the magnetic structure and performance through the liquid-liquid phase transition., Comment: number of pages is 31 and number of figures is 14, including the Supplementary Material

Out-of-distribution (OOD) detection empowers the model trained on the closed set to identify unknown data in the open world. Though many prior techniques have yielded considerable improvements, two crucial obstacles still remain. Firstly, a unified perspective has yet to be presented to view the developed arts with individual designs, which is vital for providing insights into the related directions. Secondly, most research focuses on the post-processing schemes of the pre-trained features while disregarding the superiority of end-to-end training, dramatically limiting the upper bound of OOD detection. To tackle these issues, we propose a general probabilistic framework to interpret many existing methods and an OOD-data-free model, namely Self-supervised Sampling for OOD Detection (SSOD), to unfold the potential of end-to-end learning. SSOD efficiently exploits natural OOD signals from the in-distribution (ID) data based on the local property of convolution. With these supervisions, it jointly optimizes the OOD detection and conventional ID classification. Extensive experiments reveal that SSOD establishes competitive state-of-the-art performance on many large-scale benchmarks, where it outperforms the most recent approaches, such as KNN, by a large margin, e.g., 48.99% to 35.52% on SUN at FPR95., Comment: arXiv admin note: substantial text overlap with arXiv:2301.06657

Supplementary Material 1

We propose the In-context Autoencoder (ICAE) for context compression in a large language model (LLM). The ICAE has two modules: a learnable encoder adapted with LoRA from an LLM for compressing a long context into a limited number of memory slots, and a fixed decoder which is the target LLM that can condition on the memory slots for various purposes. We first pretrain the ICAE using both autoencoding and language modeling objectives on massive text data, enabling it to generate memory slots that accurately and comprehensively represent the original context. Then, we fine-tune the pretrained ICAE on a small amount of instruct data to enhance its interaction with various prompts for producing desirable responses. Our experimental results demonstrate that the ICAE learned with our proposed pretraining and fine-tuning paradigm can effectively produce memory slots with $4\times$ context compression, which can be well conditioned on by the target LLM to respond to various prompts. The promising results demonstrate significant implications of the ICAE for its novel approach to the long context problem and its potential to reduce computation and memory overheads for LLM inference in practice, suggesting further research effort in context management for an LLM. Our code and data will be released shortly., Comment: Work in progress

Supplementary Material 2

Supplemental Material for Abdominal Penetrating Aortic Ulcer and Endovascular Treatments: A Systematic Review and Meta-Analysis by Bing Wang, Chenyang Qiu, Xinyu Yu, Yangyan He, Tianchi Chen, Ruoran Lin, Xiaohui Wang, Jun Pan, Xun Wang, Hunkun Zhang, Ziheng Wu in Vascular and Endovascular Surgery.

Supplementary Material 3

The goal of Few-Shot Continual Learning (FSCL) is to incrementally learn novel tasks with limited labeled samples and preserve previous capabilities simultaneously, while current FSCL methods are all for the class-incremental purpose. Moreover, the evaluation of FSCL solutions is only the cumulative performance of all encountered tasks, but there is no work on exploring the domain generalization ability. Domain generalization is a challenging yet practical task that aims to generalize beyond training domains. In this paper, we set up a Generalized FSCL (GFSCL) protocol involving both class- and domain-incremental situations together with the domain generalization assessment. Firstly, two benchmark datasets and protocols are newly arranged, and detailed baselines are provided for this unexplored configuration. We find that common continual learning methods have poor generalization ability on unseen domains and cannot better cope with the catastrophic forgetting issue in cross-incremental tasks. In this way, we further propose a rehearsal-free framework based on Vision Transformer (ViT) named Contrastive Mixture of Adapters (CMoA). Due to different optimization targets of class increment and domain increment, the CMoA contains two parts: (1) For the class-incremental issue, the Mixture of Adapters (MoA) module is incorporated into ViT, then cosine similarity regularization and the dynamic weighting are designed to make each adapter learn specific knowledge and concentrate on particular classes. (2) For the domain-related issues and domain-invariant representation learning, we alleviate the inner-class variation by prototype-calibrated contrastive learning. The codes and protocols are available at https://github.com/yawencui/CMoA., Comment: Submitted to International Journal of Computer Vision (IJCV)

Enhancing word usage is a desired feature for writing assistance. To further advance research in this area, this paper introduces "Smart Word Suggestions" (SWS) task and benchmark. Unlike other works, SWS emphasizes end-to-end evaluation and presents a more realistic writing assistance scenario. This task involves identifying words or phrases that require improvement and providing substitution suggestions. The benchmark includes human-labeled data for testing, a large distantly supervised dataset for training, and the framework for evaluation. The test data includes 1,000 sentences written by English learners, accompanied by over 16,000 substitution suggestions annotated by 10 native speakers. The training dataset comprises over 3.7 million sentences and 12.7 million suggestions generated through rules. Our experiments with seven baselines demonstrate that SWS is a challenging task. Based on experimental analysis, we suggest potential directions for future research on SWS. The dataset and related codes is available at https://github.com/microsoft/SmartWordSuggestions., Comment: Accepted by Findings of ACL23

Additional file 2. Supplementary Figures.

Additional file 1. Supplementary Tables.

The chemical looping steam reforming (CLSR) of biomass tar enables the process intensification for the co-preparation of syngas and high purity hydrogen. The practical application of brownmillerite-structured Ca2Fe2O5 is hindered by activity-related issues such as low fuel conversion and oxygen transfer capacity. Here, the doping of heteroatoms, e.g. Ni induces structural changes to the brownmillerite lattice, transforming it from a Pnma phase to a Pcmn one, with increased distortion of the FeO6 octahedra. The structural changes lead to the upwards shifts of the O 2p band of oxygen carrier, and subsequently improved lattice oxygen activity as well as oxygen transfer capacity. The formation of oxygen vacancy is a rate determining step during CLSR, while the Ni-doped Ca2Fe2O5 reduces the energy of oxygen vacancy formation and energy barrier for lattice oxygen migration through the bulk. During CLSR, Ca2Ni0.25Fe1.75O5 lead to significant improvement in syngas productivity, hydrogen purity and fuel conversion. Nanyang Technological University National Research Foundation (NRF) Submitted/Accepted version The research received financial support by National Natural Science Foundation of China (No. 22109048 and No. 22005112), China Postdoctoral Science Foundation (No. 2021M691121 and No. 2022T150229), and the Fundamental Research Funds for the Central Universities (No. 2021XXJS040). TX and WL acknowledge financial support by National Research Foundation under its Campus for Research Excellence and Technological Enterprise (CREATE) programme. WL also thanks Nanyang Technological University for its Start-up Grant.

Additional file 3.

Purpose Lipid nanoparticles (LNPs) are widely utilized as means to deliver mRNA molecules. However, metric connections between biodistribution and pharmacokinetics (PK) of the nanoparticle carrier and transgene expression dynamics remain largely unknown. Methods LNPs containing mRNAs encoding the firefly luciferase gene were prepared with varying sizes. Biodistributions of injected LNPs in mice were measured by fluorescence bioimaging or liquid chromatography with tandem mass spectrometry. In addition, luciferase expression levels were determined by bioluminescence imaging and enzyme activity assays. Results Some intramuscularly injected LNPs were found circulating in the system, resulting in accumulation in the liver and spleen, especially when the LNP sizes were relatively small. Bigger LNPs were more likely to remain at the injection site. Transgene expression in the liver was found most prominent compared with other organs and tissues. Conclusions Biomolecules such as mRNAs encapsulated in locally injected LNPs can reach other organs and tissues via systemic circulation. Gene expression levels are affected by the LNP biodistribution and pharmacokinetics (PK), which are further influenced by the particle size and injection route. As transfection efficiency varies in different organs, the LNP exposure and mRNA expression are not linearly correlated.

We introduce \textsc{PoliteRewrite} -- a dataset for polite language rewrite which is a novel sentence rewrite task. Compared with previous text style transfer tasks that can be mostly addressed by slight token- or phrase-level edits, polite language rewrite requires deep understanding and extensive sentence-level edits over an offensive and impolite sentence to deliver the same message euphemistically and politely, which is more challenging -- not only for NLP models but also for human annotators to rewrite with effort. To alleviate the human effort for efficient annotation, we first propose a novel annotation paradigm by a collaboration of human annotators and GPT-3.5 to annotate \textsc{PoliteRewrite}. The released dataset has 10K polite sentence rewrites annotated collaboratively by GPT-3.5 and human, which can be used as gold standard for training, validation and test; and 100K high-quality polite sentence rewrites by GPT-3.5 without human review. We wish this work (The dataset (10K+100K) will be released soon) could contribute to the research on more challenging sentence rewrite, and provoke more thought in future on resource annotation paradigm with the help of the large-scaled pretrained models.

This paper targets unsupervised skeleton-based action representation learning and proposes a new Hierarchical Contrast (HiCo) framework. Different from the existing contrastive-based solutions that typically represent an input skeleton sequence into instance-level features and perform contrast holistically, our proposed HiCo represents the input into multiple-level features and performs contrast in a hierarchical manner. Specifically, given a human skeleton sequence, we represent it into multiple feature vectors of different granularities from both temporal and spatial domains via sequence-to-sequence (S2S) encoders and unified downsampling modules. Besides, the hierarchical contrast is conducted in terms of four levels: instance level, domain level, clip level, and part level. Moreover, HiCo is orthogonal to the S2S encoder, which allows us to flexibly embrace state-of-the-art S2S encoders. Extensive experiments on four datasets, i.e., NTU-60, NTU-120, PKU-MMD I and II, show that HiCo achieves a new state-of-the-art for unsupervised skeleton-based action representation learning in two downstream tasks including action recognition and retrieval, and its learned action representation is of good transferability. Besides, we also show that our framework is effective for semi-supervised skeleton-based action recognition. Our code is available at https://github.com/HuiGuanLab/HiCo., Accepted by AAAI 2023. The code is available at http://github.com/HuiGuanLab/HiCo

We propose eXtensible Prompt (X-Prompt) for prompting a large language model (LLM) beyond natural language (NL). X-Prompt instructs an LLM with not only NL but also an extensible vocabulary of imaginary words that are introduced to help represent what NL words hardly describe, allowing a prompt to be more descriptive. Like NL prompts, X-Prompt is out-of-distribution (OOD) robust, for which we propose context-guided learning with prompt augmentation to learn its imaginary words for general usability, enabling them to use in different prompt contexts for fine-grain specifications. The promising results of X-Prompt demonstrate its potential of approaching advanced interaction between humans and LLMs to bridge their communication gap., Work in progress

Prompts with different control signals (e.g., length, keywords, etc.) can be used to control text summarization. When control signals are available, they can control the properties of generated summaries and potentially improve summarization quality (since more information are given). Unfortunately, control signals are not already available during inference time. In this paper, we propose Lotus (shorthand for Latent Prompt Tuning for Summarization), which is a single model that can be applied in both controlled and uncontrolled (without control signals) modes. During training, Lotus learns latent prompt representations from prompts with gold control signals using a contrastive learning objective. Experiments show Lotus in uncontrolled mode consistently improves upon strong (uncontrollable) summarization models across four different summarization datasets. We also demonstrate generated summaries can be controlled using prompts with user specified control tokens.

We design statistical hypothesis tests for performing leak detection in water pipeline channels. By applying an appropriate model for signal propagation, we show that the detection problem becomes one of distinguishing signal from noise, with the noise being described by a multivariate Gaussian distribution with unknown covariance matrix. We first design a test procedure based on the generalized likelihood ratio test, which we show through simulations to offer appreciable leak detection performance gain over conventional approaches designed in an analogous context (for radar detection). Our proposed method requires estimation of the noise covariance matrix, which can become inaccurate under high-dimensional settings, and when the measurement data is scarce. To deal with this, we present a second leak detection method, which employs a regularized covariance matrix estimate. The regularization parameter is optimized for the leak detection application by applying results from large dimensional random matrix theory. This second proposed approach is shown to yield improved performance in leak detection compared with the first approach, at the expense of requiring higher computational complexity.

Cadmium (Cd) contamination of paddy soil affects safe crop production. This study aimed to evaluate the effects of plant biostimulant amino acid fertilizer on the phytoremediation capability of an emergent accumulator plant Nasturtium officinale R. Br. for Cd-contaminated paddy soils. A pot study was carried out to study the effects of different concentrations of amino acid fertilizer on the Cd accumulation of N. officinale grown in Cd-contaminated paddy soil. The amino acid fertilizer increased the biomass of N. officinale. The amino acid fertilizer concentration exhibited a quadratic polynomial regression relationship with the root and shoot biomass. The fertilizer also increased the photosynthetic pigment (chlorophyll and carotenoid) contents, peroxidase (POD; EC 1.11.1.7) activity, and catalase (CAT; EC 1.11.1.6) activity of N. officinale, but decreased the soluble protein content and had no significant effect on the superoxide dismutase (SOD; EC 1.15.1.1) activity. Furthermore, the amino acid fertilizer increased the Cd content and Cd extraction of N. officinale. The shoot Cd extraction increased by 29.06%, 63.05%, 77.22%, and 17.40% at 1500-, 1200-, 900-, and 600-fold dilutions of the amino acid fertilizer, respectively, compared with the control. Moreover, the amino acid fertilizer promoted the Cd transport from the roots to shoots of N. officinale. The amino acid fertilizer concentration also exhibited a quadratic polynomial regression relationship with the root Cd content, shoot Cd content, root Cd extraction, and shoot Cd extraction, respectively. The correlation, grey relational, and path analyses revealed that the root biomass, shoot biomass, chlorophyll content, catalase activity, shoot Cd content, and root Cd extraction were closely associated with the shoot Cd extraction. Therefore, the amino acid fertilizer can promote Cd uptake and improve the phytoremediation capability of N. officinale to remediate Cd-contaminated paddy soils, and 900-fold dilution is the most suitable concentration.

BackgroundThe left atrial low-voltage areas (LVAs) are associated with atrial fibrosis; however, it is not clear how the left atrial LVAs affect the recurrence of arrhythmias after catheter ablation, and the efficacy and safety of the left atrial substrate modification based on LVAs as a strategy for catheter ablation of atrial fibrillation (AF) are not evident for AF patients with LVAs.MethodsWe performed a systematic search to compare the arrhythmia recurrence in AF patients with and without LVAs after conventional ablation and arrhythmia recurrence in LVAs patients after conventional ablation with and without substrate modification based on LVAs.ResultA total of 6 studies were included, involving 1,175 patients. The arrhythmia recurrence was higher in LVA patients after conventional ablation (OR: 5.14, 95% CI: [3.11, 8.49]; P < 0.00001). Additional LVAs substrate modification could improve the freedom of arrhythmia in LVAs patients after the first procedure (OR: 0.30, 95% CI: [0.15, 0.62]; P = 0.0009). However, there was no significant difference after multiple procedures (P = 0.19). The procedure time (MD: 26.61, 95% CI [15.79, 37.42]; P < 0.00001) and fluoroscopy time (MD: 6.90, 95% CI [4.34, 9.47]; P < 0.00001) in LVAs patients with additional LVAs substrate modification were significantly increased compared to LVAs patients’ without substrate modification. Nevertheless, there were no higher LVAs substrate modification-related complications (P = 0.93) between LVAs patients with and without additional LVAs substrate modification. In the subgroup analysis, the additional LVAs substrate modification reduced the risk of arrhythmia recurrence in LVAs patients during the follow-up time, which was 12 months (OR: 0.32, 95% CI (0.17, 0.58); P = 0.002), and box isolation (OR: 0.37, 95% CI (0.20, 0.69); P = 0.002) subgroups, but the type of AF, follow up >12 months and homogenization subgroups were not statistically significant. Trial sequential analysis shows conclusive evidence for the LVAs ablation.ConclusionThis study has shown that LVAs could improve the risk of arrhythmia recurrence in AF patients after conventional ablation. And additional LVAs substrate modification after conventional ablation could increase the freedom of arrhythmia recurrence in LVAs patients. Interestingly, the box isolation approach appeared more promising.Systematic review registration[http://www.crd.york.ac.uk/prospero], identifier [CRD42021239277].

Smart contract, one of the most successful applications of blockchain, is taking the world by storm, playing an essential role in the blockchain ecosystem. However, frequent smart contract security incidents not only result in tremendous economic losses but also destroy the blockchain-based credit system. The security and reliability of smart contracts thus gain extensive attention from researchers worldwide. In this survey, we first summarize the common types and typical cases of smart contract vulnerabilities from three levels, i.e., Solidity code layer, EVM execution layer, and Block dependency layer. Further, we review the research progress of smart contract vulnerability detection and classify existing counterparts into five categories, i.e., formal verification, symbolic execution, fuzzing detection, intermediate representation, and deep learning. Empirically, we take 300 real-world smart contracts deployed on Ethereum as the test samples and compare the representative methods in terms of accuracy, F1-Score, and average detection time. Finally, we discuss the challenges in the field of smart contract vulnerability detection and combine with the deep learning technology to look forward to future research directions., This manuscript is the English translation version of our paper published in Ruan Jian Xue Bao/Journal of Software, 22, 33(8)

Current methods for text-to-video retrieval (T2VR) are trained and tested on video-captioning oriented datasets such as MSVD, MSR-VTT and VATEX. A key property of these datasets is that videos are assumed to be temporally pre-trimmed with short duration, whilst the provided captions well describe the gist of the video content. Consequently, for a given paired video and caption, the video is supposed to be fully relevant to the caption. In reality, however, as queries are not known a priori, pre-trimmed video clips may not contain sufficient content to fully meet the query. This suggests a gap between the literature and the real world. To fill the gap, we propose in this paper a novel T2VR subtask termed Partially Relevant Video Retrieval (PRVR). An untrimmed video is considered to be partially relevant w.r.t. a given textual query if it contains a moment relevant to the query. PRVR aims to retrieve such partially relevant videos from a large collection of untrimmed videos. PRVR differs from single video moment retrieval and video corpus moment retrieval, as the latter two are to retrieve moments rather than untrimmed videos. We formulate PRVR as a multiple instance learning (MIL) problem, where a video is simultaneously viewed as a bag of video clips and a bag of video frames. Clips and frames represent video content at different time scales. We propose a Multi-Scale Similarity Learning (MS-SL) network that jointly learns clip-scale and frame-scale similarities for PRVR. Extensive experiments on three datasets (TVR, ActivityNet Captions, and Charades-STA) demonstrate the viability of the proposed method. We also show that our method can be used for improving video corpus moment retrieval., Accepted by ACM MM 2022. The paper's homepage is http://danieljf24.github.io/prvr

The seed coat takes an important function in the life cycle of plants, especially seed growth and development. It promotes the accumulation of nutrients inside the seed and protects the seed embryo from mechanical damage. Seed coat permeability is an important characteristic of seeds, which not only affects seed germination, but also hinders the detection of seed vigor by electrical conductivity (EC) method. This research aimed to elucidate the mechanism of seed coat permeability formation through metabolome and transcriptome analysis of Elymus nutans. We collected the samples at 8, 18, and 28 days post-anthesis (dpa), and conducted a seed inclusion exosmosis experiment and observed the seed coat permeability. Moreover, we analyzed the changes in the metabolome and transcriptome during different development stages. Here, taking 8 dpa as control, 252 upregulated and 157 downregulated differentially expressed metabolites (DEMs) were observed and 886 upregulated unigenes and 1170 downregulated unigenes were identified at 18 dpa, while 4907 upregulated unigenes and 8561 downregulated unigenes were identified at 28 dpa. Meanwhile, we observed the components of ABC transporters, the biosynthesis of unsaturated fatty acids, and phenylalanine metabolism pathways. The key metabolites and genes affecting seed coat permeability were thiamine and salicylic acid. Furthermore, there were 13 and 14 genes with correlation coefficients greater than 0.8 with two key metabolites, respectively, and the —log2Fold Change— of these genes were greater than 1 at different development stages. Meanwhile, pathogenesis-related protein 1 and phenylalanine ammonia-lyase play an important role in regulating the formation of compounds. Our results outline a framework for understanding the development changes during seed growth of E. nutans and provide insights into the traits of seed coat permeability and supply a great significance value to seed production and quality evaluation.

Video search has become the main routine for users to discover videos relevant to a text query on large short-video sharing platforms. During training a query-video bi-encoder model using online search logs, we identify a modality bias phenomenon that the video encoder almost entirely relies on text matching, neglecting other modalities of the videos such as vision, audio. This modality imbalanceresults from a) modality gap: the relevance between a query and a video text is much easier to learn as the query is also a piece of text, with the same modality as the video text; b) data bias: most training samples can be solved solely by text matching. Here we share our practices to improve the first retrieval stage including our solution for the modality imbalance issue. We propose MBVR (short for Modality Balanced Video Retrieval) with two key components: manually generated modality-shuffled (MS) samples and a dynamic margin (DM) based on visual relevance. They can encourage the video encoder to pay balanced attentions to each modality. Through extensive experiments on a real world dataset, we show empirically that our method is both effective and efficient in solving modality bias problem. We have also deployed our MBVR in a large video platform and observed statistically significant boost over a highly optimized baseline in an A/B test and manual GSB evaluations., Accepted by SIGIR-2022, short paper

Novel dendritic monomers derived from cystamine (Cy) were firstly synthesized in this research. Then, these monomers were used as crosslinkers to prepare three series of thermosensitive hydrogels with 2-(2-methoxyethoxy) ethyl methacrylate (MEMA). The effects of these crosslinkers on the hydrogel’s properties such as swelling behavior, mechanical properties, decomposability and thermoreversibility were investigated and compared with the hydrogel crosslinked by N,N′-bisacrylyl cystamine. The results showed that the equilibrium swelling ratio (SReq) of the hydrogel was affected by the hydrophilic group on the structure of the crosslinker and branch distribution of dendrimer crosslinker. More hydrophilic groups led to higher SReq; however, more complex branch distribution occupied the storage space of water and thus reduced the SReq. The prepared hydrogels all showed apparent thermoreversibility, and the critical gel transition temperature (CGTT) of these hydrogels increased with an increase of hydrophilic group in the crosslinker. Their CGTT showed a range from 26 °C (pure polymer) to 45 °C. More importantly, disulfide bonds in the synthesized hydrogel could be decomposed in the presence of cysteine. This result was confirmed by the change of swelling ratio and mechanical properties before and after they were treated with cysteine. The hydrogels decomposed by cysteine showed higher swelling ratio than the pristine hydrogels. And for mechanical properties, the shear modulus, Young’s modulus and effective crosslinking density of the decomposed hydrogels are lower than the pristine hydrogels. In summary, a thermoresponsive and cysteine-responsive hydrogel has been developed.

Physical geography is a basic research subject of natural sciences. Its research object is the natural environment which is closely related to human living and development, and China’s natural environment is complex and diverse. According to national needs and regional development, physical geographers have achieved remarkable achievements in applied basis and applied research, which also has substantially contributed to the planning of national economic growth and social development, the protection of macro ecosystems and resources, and sustainable regional development. This study summarized the practice and application of physical geography in China over the past 70 years in the following fields: regional differences in natural environments and physical regionalization; land use and land cover changes; natural hazards and risk reduction; process and prevention of desertification; upgrading of medium- and low-yield fields in the Huang-Huai-Hai region; engineering construction in permafrost areas; geochemical element anomalies and the prevention and control of endemic diseases; positioning and observation of physical geographical elements; and identification of geospatial differentiation and geographical detectors. Furthermore, we have proposed the future direction of applied research in the field of physical geography.

This paper details the development and application of a novel pressure-tight sampler with a metal seal capable of acquiring high-purity fluid samples from deep-sea hydrothermal vents. The sampler has a titanium diaphragm valve for sampling and a flexible titanium foil bag to store the fluid sample. Hence, all parts of the sampler in contact with the sample are made of titanium without elastomer O-ring seals to minimize the organic carbon blank of the sampler, which makes it suitable for collecting organic samples. A pressure-tight structure was specially designed to maintain the sample at in situ pressure during the recovery of the sampler. The sampler has been successfully tested in a sea trial from November 2018 to March 2019, and pressure-tight hydrothermal fluid samples have been collected.

Background: Hepatocellular carcinoma is one of the most common malignant tumors found all over the globe. Despite advances in surgery and chemotherapy, the five-year survival rate of patients with hepatocellular carcinoma is still low. It is known that the proliferation of hepatocellular carcinoma cells is closely related to the occurrence, development and prog- nosis of hepatocellular carcinoma. The present work investigates the expression of microRNA-489 (miR-489) in human hepatocellular carcinoma cells and its effect on the biological behavior of human hepatocellular carcinoma cells. Methods: The expression of miR-489 by fluorescence quantitative PCR detection in 30 patients with hepatoblastoma of liver cancer tissues and adjacent tissues was studied. Also, the determination of hepatoblastoma in four cell lines with differ- ent metastatic potential (HR8348, HCT116, HT29 and HEPG2) and the expression of miR-489 during miR-489 simulation process was studied. MTT assay, flow cytometry and Western blot analysis were performed to know the cell proliferation to detect the changes in cell cycle, apoptosis of cells, and SOX4 gene expression respectively. Results: RT-PCR results showed that the cells compared with pre-cancerous tissue, the expression level of miR-489 in hepatocellular carcinoma tissues than in adjacent tissue significantly decreased (P 0.05); but miR-489 expression could induce apoptosis, compared with the control group, the apoptosis of miR-489 mimics was significantly increased and the difference was statistically significant (P < 0.05). Conclusion: In conclusion, miR-489 can significantly inhibit the occurrence and development of hepatocellular carcinoma cells. The mechanism may be down regulated by the expression of SOX4 and inhibit cell proliferation. Further this study showed that the tumor cells SOX4 gene as a regulatory factor target the genes of miR-489 in hepatocellular carcinoma. Keywords: Hepatocellular carcinoma; mircroRNA-489; SOX4; apoptosis.

Due to cancer heterogeneity, only some patients can benefit from drug therapy. The personalized drug usage is important for improving the treatment response rate of cancer patients. The value of the transcriptome of patients has been recently demonstrated in guiding personalized drug use, and the Connectivity Map (CMAP) is a reliable computational approach for drug recommendation. However, there is still no personalized drug recommendation tool based on transcriptomic profiles of patients and CMAP. To fill this gap, here, we proposed such a feasible workflow and a user-friendly R package—Cancer-Personalized Drug Recommendation (CPDR). CPDR has three features. 1) It identifies the individual disease signature by using the patient subgroup with transcriptomic profiles similar to those of the input patient. 2) Transcriptomic profile purification is supported for the subgroup with high infiltration of non-cancerous cells. 3) It supports in silico drug efficacy assessment using drug sensitivity data on cancer cell lines. We demonstrated the workflow of CPDR with the aid of a colorectal cancer dataset from GEO and performed the in silico validation of drug efficacy. We further assessed the performance of CPDR by a pancreatic cancer dataset with clinical response to gemcitabine. The results showed that CPDR can recommend promising therapeutic agents for the individual patient. The CPDR R package is available at https://github.com/AllenSpike/CPDR.

The multicopy suppressor of IRA (MSI) is a subfamily of WD40 repeat proteins, which is widely involved in plant growth and development. In order to explore the function of MdMSI members in abiotic stress, we identified eight MSI gene family members from the Malus × domestica reference genome. They were distributed on six chromosomes, and they had similar secondary and tertiary structures. We found a variety of regulatory elements in response to hormones and abiotic stress in MdMSI promoters. Through qRT-PCR analysis, it was revealed that MdMSIs were expressed in all tissues, especially in roots. The analysis results also revealed that the expression of MdMSIs was induced in varying degrees under salt, drought stress, and ABA treatments. Furthermore, we obtained the overexpression of MdMSI1-1 transgenic apple calli and Arabidopsis. The overexpression of MdMSI1-1 in calli and Arabidopsis played a negative regulatory role in salt stress response. Our work laid a foundation for further verifying the function of MSI genes under abiotic stress in apples.

Existing field potato grading devices are complex in structure and large in volume, designing a rod-type conveyor grading device. Relying on field measurement, size and field distribution laws of mature potatoes are obtained and their field distribution is modelled. Moreover, the conveyor grading device model is established using the discrete element method (EDEM). The single-factor test and multi-factor test adopting one evaluation index (grading accuracy) and four test factors (conveyor chain elevation angle, conveyor chain speed, rod clearance, and potato feed rate) were conducted. Finally, the effects of various test factors on the evaluation index are analyzed and test factors are optimized. According to the results of this study, the rod-type conveyor grading device had a grading accuracy of 93.32%. The relative error between experimental validation and simulation results was less than 8%, suggesting that the regressive mathematical model proposed in this paper and the optimization results obtained were reasonable.