Search

Ding, Yipu, Liu, Zinuan, Wang, Xi, Xin, Ran, Shan, Dongkai, He, Bai, Jing, Jing, Gao, Qi, Yang, Junjie, Chen, Yundai (2023): Validation of hemodynamic stress calculation in coronary computed tomography angiography versus intravascular ultrasound. Journal of Herpetology 13 (4): 298-309, DOI: 10.1670/20-035, URL: http://dx.doi.org/10.21037/qims-22-832

Thermodynamics of glassy states in a quasi-two-dimensional frustrated magnet Ba$_2$Sn$_2$ZnCr$_{7p}$Ga$_{10-7p}$O$_{22}$ where $p$ is the spin density are investigated experimentally. The system features a triangular network of bipyramids of spins with the quantum spin number $s = 3/2$. The DC magnetic susceptibility measurements on a series of samples with $0.44 \le p \le 0.98$ show a freezing transition with the transition temperature $T_\mathrm{f} \le 1.2$ K. $T_\mathrm{f}$ is found to decrease with decreasing $p$. The low-lying excitations in the glassy state of the system are examined via the temperature dependence of the magnetic heat capacity and are shown to consist of two components: the hydrodynamic Halperin-Saslow modes characteristic of a spin jam and the two-level systems of a spin glass. A continuous crossover between the two glassy states is observed via the varying weights of the two components as the spin density is varied. The $p$ dependence of the spin jam's zero-point entropy determined from the exotic perimeter-scaling behavior combined with the observed zero-point entropy of the samples provides the $p$ dependence of the spin glass's zero-point entropy. The obtained result shows that the correlations between orphan spins begin below $p \sim 0.8$, the limit that was also found using a neutron scattering technique in a previous report on the isostructural compound SrCr$_{9p}$Ga$_{12-9p}$O$_{19}$. The domain size of the spin-jam state estimated from the value of the zero-point entropy for the cleanest sample is approximately $4 \times 4$ bipyramids, about 2.5 times the measured spin correlation length.

Learning to optimize (L2O) has gained increasing popularity, which automates the design of optimizers by data-driven approaches. However, current L2O methods often suffer from poor generalization performance in at least two folds: (i) applying the L2O-learned optimizer to unseen optimizees, in terms of lowering their loss function values (optimizer generalization, or ``generalizable learning of optimizers"); and (ii) the test performance of an optimizee (itself as a machine learning model), trained by the optimizer, in terms of the accuracy over unseen data (optimizee generalization, or ``learning to generalize"). While the optimizer generalization has been recently studied, the optimizee generalization (or learning to generalize) has not been rigorously studied in the L2O context, which is the aim of this paper. We first theoretically establish an implicit connection between the local entropy and the Hessian, and hence unify their roles in the handcrafted design of generalizable optimizers as equivalent metrics of the landscape flatness of loss functions. We then propose to incorporate these two metrics as flatness-aware regularizers into the L2O framework in order to meta-train optimizers to learn to generalize, and theoretically show that such generalization ability can be learned during the L2O meta-training process and then transformed to the optimizee loss function. Extensive experiments consistently validate the effectiveness of our proposals with substantially improved generalization on multiple sophisticated L2O models and diverse optimizees. Our code is available at: https://github.com/VITA-Group/Open-L2O/tree/main/Model_Free_L2O/L2O-Entropy., This paper is accepted in AISTATS 2023

Objective: To assess the clinical and prognostic value of MRI perfusion-weighted imaging (PWI) versus diffusion-weighted imaging (DWI) parameters for apparent diffusion coefficient (ADC) and cerebral blood flow (CBF) in the diagnosis of patients with ischemic stroke (IS). Methods: Eighty patients diagnosed with IS in the Second Affiliated Hospital of Soochow University from January 2020 to January 2021 were retrospectively analyzed and regarded as a patient group. Meanwhile, 50 patients who underwent physical examination at the Second Affiliated Hospital of Soochow University during the same period were collected, and were identified by physical examination to have atherosclerotic stenosis but not cerebral infarction, they were set as a control group. The differences of ADC and CBF between both groups were compared. The diagnostic value of ADC and CBF in diagnosing acute ischemic stroke was analyzed by receiver operating characteristic (ROC) curve. The changes of ADC and CBF before and after treatment were compared. Patients were sub-grouped according to their mRS scores, and those with scores of 0-2 were grouped into the good prognosis group while those with scores of 3-6 were grouped into the poor prognosis group, and the risk factors affecting patients’ prognosis were evaluated by logistic regression. The correlation of ADC and CBF with National Institutes of Health Stroke Scale (NIHSS) scores and modified Rankin Scale (mRS) scores was analyzed. ADC and CBF levels were compared between deceased and surviving patients, and their predictive value was assessed by ROC curves. Results: ADC and CBF were dramatically lower in the patient group compared with the control group (P

Learning to Optimize (L2O) has drawn increasing attention as it often remarkably accelerates the optimization procedure of complex tasks by ``overfitting" specific task type, leading to enhanced performance compared to analytical optimizers. Generally, L2O develops a parameterized optimization method (i.e., ``optimizer") by learning from solving sample problems. This data-driven procedure yields L2O that can efficiently solve problems similar to those seen in training, that is, drawn from the same ``task distribution". However, such learned optimizers often struggle when new test problems come with a substantially deviation from the training task distribution. This paper investigates a potential solution to this open challenge, by meta-training an L2O optimizer that can perform fast test-time self-adaptation to an out-of-distribution task, in only a few steps. We theoretically characterize the generalization of L2O, and further show that our proposed framework (termed as M-L2O) provably facilitates rapid task adaptation by locating well-adapted initial points for the optimizer weight. Empirical observations on several classic tasks like LASSO and Quadratic, demonstrate that M-L2O converges significantly faster than vanilla L2O with only $5$ steps of adaptation, echoing our theoretical results. Codes are available in https://github.com/VITA-Group/M-L2O., Comment: This paper is accepted in ICLR 2023

Using a systematic ab initio quantum many-body approach that goes beyond low-energy models, we directly compute the superconducting pairing order of several doped cuprate materials and structures. We find that we can correctly capture two well-known trends: the pressure effect, where pairing order increases with intra-layer pressure, and the layer effect, where the pairing order varies with the number of copper-oxygen layers. From these calculations, we observe that the strength of superexchange and the covalency at optimal doping are the best descriptors of the maximal pairing order. Our microscopic analysis further identifies short-range copper spin fluctuations, together with multi-orbital charge fluctuations, as central to the pairing trends. Our work illustrates the possibility of a quantitative computational understanding of unconventional high-temperature superconducting materials., Comment: 10 pages, 5 figures, with supplementary materials

Cancer immunotherapy has drawn much attention because it can restart the recognition and killing function of the immune system to normalize the antitumor immune response. However, the role of radiotherapy and chemotherapy in cancer treatment cannot be ignored. Due to cancer heterogeneity, combined therapy has become a new trend, and its efficacy has been confirmed in many studies. This review discussed the clinical implications and the underlying mechanisms of cancer immunotherapy in combination with radiotherapy or chemotherapy, offering an outline for clinicians as well as inspiration for future research.

Machine reading comprehension (MRC) is an AI challenge that requires machine to determine the correct answers to questions based on a given passage. MRC systems must not only answer question when necessary but also distinguish when no answer is available according to the given passage and then tactfully abstain from answering. When unanswerable questions are involved in the MRC task, an essential verification module called verifier is especially required in addition to the encoder, though the latest practice on MRC modeling still most benefits from adopting well pre-trained language models as the encoder block by only focusing on the "reading". This paper devotes itself to exploring better verifier design for the MRC task with unanswerable questions. Inspired by how humans solve reading comprehension questions, we proposed a retrospective reader (Retro-Reader) that integrates two stages of reading and verification strategies: 1) sketchy reading that briefly investigates the overall interactions of passage and question, and yield an initial judgment; 2) intensive reading that verifies the answer and gives the final prediction. The proposed reader is evaluated on two benchmark MRC challenge datasets SQuAD2.0 and NewsQA, achieving new state-of-the-art results. Significance tests show that our model is significantly better than the strong ELECTRA and ALBERT baselines. A series of analysis is also conducted to interpret the effectiveness of the proposed reader., Accepted by AAAI 2021

Yuhong Jin,1 Haiyan Mao,1 Bingyang Liu,1 Fen Zhou,1 Junjie Yang,1 Lei Xu,1 Jingtao Tong,2 Chen Huang,3 Yi Ding1 1Department of Intensive Care, Lihuili Hospital, Ningbo Medical Center, Ningbo, People’s Republic of China; 2Department of Radiotherapy, Lihuili Hospital, Ningbo Medical Center, Ningbo, People’s Republic of China; 3Department of Respiratory Medicine, Lihuili Hospital, Ningbo Medical Center, Ningbo, People’s Republic of ChinaCorrespondence: Chen HuangDepartment of Respiratory Medicine, Lihuili Hospital, Ningbo Medical Center, Ningbo 315000, People’s Republic of ChinaTel +86-574-87018701Fax +86- 574-87392232Email drhchen@163.comYi DingDepartment of Intensive Care, Lihuili Hospital, Ningbo Medical Center, Ningbo 315000, People’s Republic of ChinaTel +86-574-87018661Email dingdingnb@163.comObjective: There is a paucity of published data to evaluate the efficacy and safety of imipenem (IPM) and piperacillin-tazobactam (PT) dosing regimens in the treatment of septic patients acquiring continuous renal replacement therapy (CRRT).Methods and Materials: Critically-ill patients were grouped into short-stay and long-stay intensive care unit (ICU) patients. Pathogens were isolated from bloodstream infections in these patients. Minimum inhibitory concentration (MIC) value was determined by agar dilution method. Population PK models were introduced in this study, and differences in the likelihood of achieving efficacious and toxic exposures of IPM and PT for critically-ill patients were assessed.Results: A total of 86 K. pneumoniae bloodstream infection associated isolates were collected, and the MIC50 and MIC90 for short-stay ICU patients were 0.5/4 mg/L and 32/128 mg/L, respectively. IMP 0.5g q8h reached 90% probability of target attainment (PTA) against isolates with MICs ≤ 2 mg/L and was recommended to empirically treat short-stay ICU patients during CRRT based on the target of 40% ƒT>MIC. However, based on a more aggressive target of 100% ƒT>MIC, all the simulated IMP regimens except for IMP 1g q6h failed to achieve > 80% cumulative fraction of response (CFR) in such patients. Unfortunately, the risk of drug-related toxicity for IMP 1g q6h was relatively high (50– 85%). For PT, even the regimen of 4/0.5g q6h failed to provide sufficient antimicrobial exposure in short-stay ICU patients acquiring CRRT.Conclusion: No dose adjustment was required for the conventional IMP and PT regimens in the critically-ill population acquiring CRRT. Empirical treatment of IMP 0.5g q8h/q6h, not for PT, may provide sufficient antimicrobial exposure for short-stay ICU patients during CRRT. PT should be used in the knowledge of MIC results.Keywords: Klebsiella pneumoniae infection, pharmacokinetic/pharmacodynamic, continuous renal replacement therapy; CRRT, intensive care unit; ICU

Objective: Rheumatoid arthritis (RA) is a chronic, progressive autoimmune disease characterized by aggressive and symmetric polyarthritis. Our previous study showed that celastrol (Cel) protected against RA by inhibiting the NLRP3 inflammasome pathway, but the molecular mechanism has not been clarified. Methods: A type II collagen-induced arthritis (CIA) DBA/1 mouse model was used to study the antiarthritic effects of Cel, and paw swelling, arthritis scores, serum cytokine levels, and pathological examinations were carried out. The effects of Cel on the fibroblast-like synoviocytes (FLSs) viability, proliferation and migration of tumour necrosis factor-α (TNF-α)-induced FLSs were tested by MMT assays, EdU staining and scratch wound healing assays. The proinflammatory factors were evaluated by ELISA. The expression of NLRP3, ASC, cleaved caspase-1, p-p65, p65 and p-IκB-α was examined in vitro and in vivo by western blotting and immunofluorescence analysis, respectively. The putative binding sites between Cel and Hsp90 were predicted through molecular docking analysis, the Octet RED96 system and coimmunoprecipitation. Results: The results showed that Cel decreased arthritis severity and reduced TNF-α-induced FLS migration and proliferation. Additionally, Cel decreased the protein expression of NLRP3, ASC, and cleaved caspase-1 and the secretion of proinflammatory cytokines. Furthermore, Cel directly interacted with Hsp90, which interacts with NLRP3, and Cel blocked the interaction between Hsp90 and NLRP3 in FLSs. Conclusion: In summary, our findings demonstrate that Cel regulates NLRP3 inflammasome pathway in vivo and in vitro. Cel inhibits the proliferation and migration of FLSs by blocking the interaction between Hsp90 and NLRP3, thus inhibiting NLRP3 inflammasome pathway.

Ultra-thin SiOx (with x ≤ 2) layer plays a key role in Si technology especially for the CMOS foundry processes approaching nm-level. Besides various functions of SiOx by only manipulating x value and thickness, [1-3] it becomes a favourable material as surface protecting layer for Si-molecular beam epitaxy (Si MBE), especially important for the growth on patterned Si substrates after a series of fabrication processes. This is because SiOx can be thermally decomposed to generate an atomically flat surface, without introducing any reducing agent (e.g., H2) to affect the growth.[4] Moreover, recent MBE growth indicates that surface reconstructions with bi-atomic terraces can form by thermally removing an oxide layer, which is critical to supress antiphase boundary defects towards high quality III-V materials monolithic grown on Si.[5] However, there has been insufficient knowledge in preparing a reliable thin SiOx protecting layer for MBE growth so far, and consistent confusions of the concepts have been used in literatures regarding the fabrication methods.In this work, we systematically studied the oxidation and deoxidation mechanisms of ultra-thin SiOx layers fabricated on Si (100) for MBE growth. From several growth approaches, a chemically synthesized SiOx thin layer is developed for reproducible thermal deoxidation in a Si MBE chamber. From transmission electron microscopy (TEM) observations, the SiOx thickness is ~2 nm with a clear Si/SiOx boundary to stop further oxidation in air (Fig.a), working well as a surface protection layer. An atomically flat surface is realized after thermal deoxidation process, examined by atomic force microscopy (AFM). The subsequent epitaxial growth is performed with the same MBE by depositing a thin Ge epitaxial layer, as it is not only broadly applied for all group IV integrations but also a good buffer layer for III-V on Si growth. [6,7] The epitaxial layer exhibits a reduced thread dislocation defect density (down to 3×108/cm-2) compared with its counterpart grown on a thermally removed natural oxide layer. In the meantime, a reduced tensile strain from 0.29% to 0.24% were confirmed by x-ray diffraction (XRD) and Raman spectroscopy measurements, corresponding to a variation of Ge-Γ1 transition energy from 0.772 to 0.783 eV confirmed by absorption spectroscopy measurements (Fig.b). Our work introduces a SiOx surface protecting layer for high quality MBE growth on Si substrates, opening a way for both all-group IV and III-V on Si integrations.

Drug addiction is a serious health problem prevalent worldwide. Currently available therapies including pharmacotherapy and psychotherapy are insufficient to meet the clinical needs for treating drug abuse. Recently, immunotherapy has emerged as a promising approach to treat such drug-use disorders. Pharmacokinetic antagonists are used in immunotherapy, functioning by sequestering the drugs in the periphery but without allowing the drug to cross the blood-brain barrier. This can reduce the toxic and rewarding effects of the drugs, while preventing addiction and facilitating reduced relapse rates. Herein, we update recent developments in the immunotherapeutic strategies to treat abuse of drugs like methamphetamine, heroin and cocaine. In addition, we summarize the drug design used so far and its optimization strategies. Further, we document the efficacy of anti-drug vaccines and monoclonal antibodies, with an aim to promote development of new anti-drug immunotherapies.

In the practical application of brain-machine interface technology, the problem often faced is the low information content and high noise of the neural signals collected by the electrode and the difficulty of decoding by the decoder, which makes it difficult for the robotic to obtain stable instructions to complete the task. The idea based on the principle of cooperative shared control can be achieved by extracting general motor commands from brain activity, while the fine details of the movement can be hosted to the robot for completion, or the brain can have complete control. This study proposes a brain-machine interface shared control system based on spiking neural networks for robotic arm movement control and wheeled robots wheel speed control and steering, respectively. The former can reliably control the robotic arm to move to the destination position, while the latter controls the wheeled robots for object tracking and map generation. The results show that the shared control based on brain-inspired intelligence can perform some typical tasks in complex environments and positively improve the fluency and ease of use of brain-machine interaction, and also demonstrate the potential of this control method in clinical applications of brain-machine interfaces., 16 pages, 16 figures

Additional file 1.

The development of artificial intelligence (AI) and robotics are both based on the tenet of "science and technology are people-oriented", and both need to achieve efficient communication with the human brain. Based on multi-disciplinary research in systems neuroscience, computer architecture, and functional organic materials, we proposed the concept of using AI to simulate the operating principles and materials of the brain in hardware to develop brain-inspired intelligence technology, and realized the preparation of neuromorphic computing devices and basic materials. We simulated neurons and neural networks in terms of material and morphology, using a variety of organic polymers as the base materials for neuroelectronic devices, for building neural interfaces as well as organic neural devices and silicon neural computational modules. We assemble organic artificial synapses with simulated neurons from silicon-based Field-Programmable Gate Array (FPGA) into organic artificial neurons, the basic components of neural networks, and later construct biological neural network models based on the interpreted neural circuits. Finally, we also discuss how to further build neuromorphic devices based on these organic artificial neurons, which have both a neural interface friendly to nervous tissue and interact with information from real biological neural networks., 12 pages, 10 figures

Animal pose estimation and tracking (APT) is a fundamental task for detecting and tracking animal keypoints from a sequence of video frames. Previous animal-related datasets focus either on animal tracking or single-frame animal pose estimation, and never on both aspects. The lack of APT datasets hinders the development and evaluation of video-based animal pose estimation and tracking methods, limiting real-world applications, e.g., understanding animal behavior in wildlife conservation. To fill this gap, we make the first step and propose APT-36K, i.e., the first large-scale benchmark for animal pose estimation and tracking. Specifically, APT-36K consists of 2,400 video clips collected and filtered from 30 animal species with 15 frames for each video, resulting in 36,000 frames in total. After manual annotation and careful double-check, high-quality keypoint and tracking annotations are provided for all the animal instances. Based on APT-36K, we benchmark several representative models on the following three tracks: (1) supervised animal pose estimation on a single frame under intra- and inter-domain transfer learning settings, (2) inter-species domain generalization test for unseen animals, and (3) animal pose estimation with animal tracking. Based on the experimental results, we gain some empirical insights and show that APT-36K provides a valuable animal pose estimation and tracking benchmark, offering new challenges and opportunities for future research. The code and dataset will be made publicly available at https://github.com/pandorgan/APT-36K., Comment: Neurips 2022 dataset and benchmark track

The prostanoid EP4 receptor is one of the key receptors associated with inflammatory mediator PGE2-elicited immunosuppression in the tumor microenvironment. Blockade of EP4 signaling to enhance immunity-mediated tumor elimination has recently emerged as a promising strategy for cancer immunotherapy. In our efforts to discover novel subtype-selective EP4 antagonists, we designed and synthesized a class of 1H-1,2,3-triazole-based ligands that display low nanomolar antagonism activity toward the human EP4 receptor and excellent subtype selectivity. The most promising compound 59 exhibits single-digit nanomolar potency in the EP4 calcium flux and cAMP-response element reporter assays and effectively suppresses the expression of multiple immunosuppression-related genes in macrophage cells. On the basis of its favorable ADMET properties, compound 59 was chosen for further in vivo biological evaluation. Oral administration of compound 59 significantly inhibited tumor growth in the mouse CT26 colon carcinoma model accompanied by enhanced infiltration of cytotoxic T lymphocytes in the tumor tissue.

Nowadays with the development of smart residential microgrid (RMG), the distributed energy storage system (DESS) can help consumers to not only balance generation and consumption but also participate in demand respond. However, the unadjustable capacity of DESS and the lack of energy sharing among users have become the major challenges to the further development of RMG. This paper proposes a novel electricity scheduling architecture based on energy cloud (EC) for RMGs and designs an electricity scheduling optimisation. The EC is used in order to link different end-users and promote coordination. In the proposed EC-based electricity scheduling architecture, the mathematical model for the RMG is provided. Moreover, considering the depreciation cost of battery, the optimisation model is established with the objective of minimising the electricity cost. Compared with the traditional RMG, simulation results show that the proposed strategy can not only allow consumers to adjust their optimal energy storage capacity but also further reduce electricity payment costs. The designed strategy provides a new and effective research perspective for electricity scheduling of RMGs.

Hydrogen sensors with fast response and recovery rate based on nanoporous palladium (Pd) and titanium dioxide (TiO2) composite films supported by anodic aluminum oxide (AAO) template have been demonstrated. Nanoporous TiO2 film was sprayed on the porous AAO templates, followed by Pd film deposited on TiO2 layer by DC magnetron sputtering. We have researched the detection performance of the hydrogen sensors depending on different thickness of TiO2 layer from 6 to 30 nm with keeping the thickness of Pd as 30 nm. The results have demonstrated the sensors with 10 nm thickness of TiO2 achieve the best performance with a response/recovery time as short as 4/8s at 0.8% and 0.4% hydrogen concentration, respectively. The sensors exhibited very good performance under hydrogen concentrations from 0.4% to 1.8%.

Immune checkpoint blockade (ICB) has a limited effect on colorectal cancer, underlining the requirement of co‐targeting the complementary mechanisms. Here, we identified prostaglandin E2 (PGE2) receptor 4 (EP4) as the master regulator of immunosuppressive myeloid cells (IMCs), which are the major driver of resistance to ICB therapy. PGE2‐bound EP4 promotes the differentiation of immunosuppressive M2 macrophages and myeloid‐derived suppressor cells (MDSCs) and reduces the expansion of immunostimulated M1 macrophages. To explore the immunotherapeutic role of EP4 signaling, we developed a novel and selective EP4 antagonist TP‐16. TP‐16 effectively blocked the function of IMCs and enhanced cytotoxic T‐cell‐mediated tumor elimination in vivo. Cell co‐culture experiments revealed that TP‐16 promoted T‐cell proliferation, which was impaired by tumor‐derived CD11b+ myeloid cells. Notably, TP‐16 and anti‐PD‐1 combination therapy significantly impeded tumor progression and prolonged mice survival. We further demonstrated that TP‐16 increased responsiveness to anti‐PD‐1 therapy in an IMC‐related spontaneous colorectal cancer mouse model. In summary, this study demonstrates that inhibition of EP4‐expressing IMCs may offer a potential strategy for enhancing the efficacy of immunotherapy for colorectal cancer., Immunosuppressive myeloid cells (IMCs) are a prominent driver of immunotherapy resistance in colorectal cancer. This study identifies EP4 as a master regulator of IMCs and highlights blockade of EP4 as a novel therapeutic strategy for enhancing immunotherapy in colorectal cancer.

Bilevel optimization has been widely applied in many important machine learning applications such as hyperparameter optimization and meta-learning. Recently, several momentum-based algorithms have been proposed to solve bilevel optimization problems faster. However, those momentum-based algorithms do not achieve provably better computational complexity than $\mathcal{\widetilde O}(\epsilon^{-2})$ of the SGD-based algorithm. In this paper, we propose two new algorithms for bilevel optimization, where the first algorithm adopts momentum-based recursive iterations, and the second algorithm adopts recursive gradient estimations in nested loops to decrease the variance. We show that both algorithms achieve the complexity of $\mathcal{\widetilde O}(\epsilon^{-1.5})$, which outperforms all existing algorithms by the order of magnitude. Our experiments validate our theoretical results and demonstrate the superior empirical performance of our algorithms in hyperparameter applications., Comment: This paper is accepted in NeurIPS 2021

Synthetic cannabinoids, as exemplified by SDB-001 (1), bind to both CB1 and CB2 receptors and exert cannabimimetic effects similar to (-)-trans-Δ9-tetrahydrocannabinol, the main psychoactive component present in the cannabis plant. As CB1 receptor ligands were found to have severe adverse psychiatric effects, increased attention was turned to exploiting the potential therapeutic value of the CB2 receptor. In our efforts to discover novel and selective CB2 receptor agonists, 1 was selected as a starting point for hit molecule identification and a class of 1H-pyrazole-3-carboxamide derivatives were thus designed, synthesized, and biologically evaluated. Systematic structure-activity relationship investigations resulted in the identification of the most promising compound 66 as a selective CB2 receptor agonist with favorable pharmacokinetic profiles. Especially, 66 treatment significantly attenuated dermal inflammation and fibrosis in a bleomycin-induced mouse model of systemic sclerosis, supporting that CB2 receptor agonists might serve as potential therapeutics for treating systemic sclerosis.

Additional file 2: Figure 2. Immunohistochemistry staining and real-time PCR for the protein and mRNA expressions of IL-1β in the fracture haematoma/callus of sham and OVX mice.

Inspired by the formulation of quantum-electrodynamical time-dependent density functional theory (QED-TDDFT) by Rubio and coworkers, we propose an implementation that uses dimensionless amplitudes for describing the photonic contributions to QED-TDDFT electron-photon eigenstates. The leads to a symmetric QED-TDDFT coupling matrix, which is expected to facilitate the future development of analytic derivatives. Through a Gaussian atomic basis implementation of the QED-TDDFT method, we examined the effect of dipole self-energy, rotating wave approximation, and the Tamm-Dancoff approximation on the QED-TDDFT eigenstates of model compounds (ethene, formaldehyde, and benzaldehyde) in an optical cavity. We highlight, in the strong coupling regime, the role of higher-energy and off-resonance excited states with large transition dipole moments in the direction of the photonic field, which are automatically accounted for in our QED-TDDFT calculations and might substantially affect the energy and composition of polaritons associated with lower-energy electronic states., Comment: 23 pages, 6 figures and 5 tables

Additional file 1: Figure 1. Radiographic changes of the tibias in sham and OVX mice at different time points of fracture healing.

Chaoshan Han,1,* Zhiwei Zhang,2,* Jiacheng Sun,1 Ke Li,3 Yangxin Li,4 Chuanlu Ren,5 Qingyou Meng,4 Junjie Yang1 1Department of Biomedical Engineering, School of Medicine and School of Engineering, University of Alabama at Birmingham, Birmingham, AL 35294, USA; 2Department of Cardiothoracic Surgery, The Second Affiliated Hospital of Soochow University, Suzhou 215006, People’s Republic of China; 3Department of Burn and Plastic Surgery, The First Affiliated Hospital of Soochow University, Suzhou 215006, People’s Republic of China; 4Department of Cardiovascular Surgery of the First Affiliated Hospital & Institute for Cardiovascular Science, Soochow University, Suzhou 215006, People’s Republic of China; 5Department of Clinical Laboratory, The 904th Hospital of the People’s Liberation Army, Wuxi 214044, People’s Republic of China*These authors contributed equally to this workCorrespondence: Junjie YangDepartment of Biomedical Engineering, School of Medicine and School of Engineering, University of Alabama at Birmingham, 1670 University Blvd, Volker Hall G094Q, Birmingham, AL 35294, USATel +1 205 934-2798Email jjyang@uab.eduQingyou MengDepartment of Cardiovascular Surgery, The First Affiliated Hospital of Soochow University, No. 899, Pinghai Road 215006 SuzhouTel +86-512-6797-2072Fax +86 -512-6778-0100Email mengqy@163.comAbstract: Ischemic diseases, especially in the heart and the brain, have become a serious threat to human health. Growth factor and cell therapy are emerging as promising therapeutic strategies; however, their retention and sustainable functions in the injured tissue are limited. Self-assembling peptide (SAP)-based hydrogels, mimicking the extracellular matrix, are therefore introduced to encapsulate and controllably release cells, cell-derived exosomes or growth factors, thus promoting angiogenesis and tissue recovery after ischemia. We will summarize the classification, composition and structure of SAPs, and the influencing factors for SAP gelation. Moreover, we will describe the functionalized SAPs, and the combinatorial therapy of cells, exosomes or growth factors with functionalized SAPs for angiogenic process as well as its advantage in immunogenicity and injectability. Finally, an outlook on future directions and challenges is provided.Keywords: self‐assembling peptide, hydrogel, angiogenesis, survival, retention

HfO2, a simple binary oxide, holds ultra-scalable ferroelectricity integrable into silicon technology. Polar orthorhombic (Pbc21) form in ultra-thin-films ascribes as the plausible root-cause of the astonishing ferroelectricity, which has thought not attainable in bulk crystals. Though, perplexities remain primarily due to the polymorphic nature and the characterization challenges at small-length scales. Herein, utilizing a state-of-the-art Laser-Diode-heated Floating Zone technique, we report ferroelectricity in bulk single-crystalline HfO2:Y as well as the presence of anti-polar Pbca phase at different Y concentrations. Neutron diffraction and atomic imaging demonstrate (anti-)polar crystallographic signatures and abundant 90o/180o ferroelectric domains in addition to the switchable polarization with little wake-up effects. Density-functional theory calculations suggest that the Yttrium doping and rapid cooling are the key factors for the desired phase. Our observations provide new insights into the polymorphic nature and phase controlling of HfO2, remove the upper size limit for ferroelectricity, and also pave a new road toward the next-generation ferroelectric devices., 34 pages. This is the original preprint before peer review. The final version has been accepted by Nature Materials

Bilevel optimization has arisen as a powerful tool for many machine learning problems such as meta-learning, hyperparameter optimization, and reinforcement learning. In this paper, we investigate the nonconvex-strongly-convex bilevel optimization problem. For deterministic bilevel optimization, we provide a comprehensive convergence rate analysis for two popular algorithms respectively based on approximate implicit differentiation (AID) and iterative differentiation (ITD). For the AID-based method, we orderwisely improve the previous convergence rate analysis due to a more practical parameter selection as well as a warm start strategy, and for the ITD-based method we establish the first theoretical convergence rate. Our analysis also provides a quantitative comparison between ITD and AID based approaches. For stochastic bilevel optimization, we propose a novel algorithm named stocBiO, which features a sample-efficient hypergradient estimator using efficient Jacobian- and Hessian-vector product computations. We provide the convergence rate guarantee for stocBiO, and show that stocBiO outperforms the best known computational complexities orderwisely with respect to the condition number $\kappa$ and the target accuracy $\epsilon$. We further validate our theoretical results and demonstrate the efficiency of bilevel optimization algorithms by the experiments on meta-learning and hyperparameter optimization., Comment: 29 pages, 19 figures and 2 tables. Published in ICML 2021

Generative machine reading comprehension (MRC) requires a model to generate well-formed answers. For this type of MRC, answer generation method is crucial to the model performance. However, generative models, which are supposed to be the right model for the task, in generally perform poorly. At the same time, single-span extraction models have been proven effective for extractive MRC, where the answer is constrained to a single span in the passage. Nevertheless, they generally suffer from generating incomplete answers or introducing redundant words when applied to the generative MRC. Thus, we extend the single-span extraction method to multi-span, proposing a new framework which enables generative MRC to be smoothly solved as multi-span extraction. Thorough experiments demonstrate that this novel approach can alleviate the dilemma between generative models and single-span models and produce answers with better-formed syntax and semantics., AAAI-21 SDU Workshop

Motivated by the search for type-II multiferroics, we present a comprehensive optical study of a complex oxide family of type-II multiferroic candidates: RbFe(MoO4)2, RbFe(SeO4)2, and RbFe(SO4)2. We employ rotational-anisotropy second harmonic generation spectroscopy (RA SHG), a technique sensitive to point symmetries, to address discrepancies in literature-assigned point/space groups and to identify the correct crystal structures. At room temperature we find that our RA SHG patterns rotate away from the crystal axes in RbFe(AO4)2 (A = Se, S), which identifies the lack of mirror symmetry and in-plane two-fold rotational symmetry. Also, the SHG efficiency of RbFe(SeO4)2 is two orders of magnitude stronger than RbFe(AO4)2 (A = Mo, S), which suggests broken inversion symmetry. Additionally, we present temperature-dependent linear optical characterizations near the band edge of this family of materials using ultraviolet-visible (UV-VIS) absorption spectroscopy. Included is experimental evidence of the band gap energy and band gap transition type for this family. Previously unreported sub-band gap absorption is also presented, which reveals prominent optical transitions, some with an unusual central energy temperature dependence. Furthermore, we find that by substituting the A-site in RbFe(AO4)2 (A = Mo, Se, S), the aforementioned transitions are spectrally tunable. Finally, we discuss the potential origin and impact of these tunable transitions., 22 pages total: 12 pages of main text, 3 pages of appendix, 1 page of references, 6 pages of figures. 7 figures total: 4 figures in the main body, 3 figures in the appendix. All figure and figure captions are listed at the end of the manuscript

E. coli BL21 (DE3) is an excellent and widely used host for recombinant protein production. Many variant hosts were developed from from BL21 (DE3), but improving the expression of specific proteins remains a major challenge in biotechnology. In this study, we found that when BL21 (DE3) overexpressed glucose dehydrogenase (GDH), a significant industrial enzyme, serious autolysis was induced. Subsequently, we observed this phenomenon in the expression of 10 other recombinant proteins. This precludes a further increase of the produced enzyme activity by extending the fermentation time, which is not conducive to the reduction of industrial enzyme production costs. The membrane structure and mRNA expression analysis showed that cells suffered programmed cell death (PCD) during autolysis period. However, blocking three known PCD pathway in BL21 (DE3) cannot alleviate autolysis completely. Furthermore, we attempted to develop a strong expression host resistant to autolysis by controlling the speed of recombinant protein expression. To find a more suitable protein expression rate, the high- and low-strength promoter lacUV5 and lac were shuffled and recombined to yield the promoter variants lacUV5-1A and lac-1G. The results showed that only one base in lac promoter needs to be changed, and the A at the +1 position was changed to a G, resulting in a host of BL21 (DE3-lac1G), which successfully withstand the PCD of the host. The GDH activity at 43h was greatly increased from 37.5 U/mL to 452.0 U/mL. In scale-up fermentation, the new host was able to produce the model enzyme with a high rate of 89.55 U/mL/h at 43h, compared to the 3 U/mL/h of BL21 (DE3). Importantly, BL21 (DE3-lac1G) also successfully improved the production of other 10 enzymes. The engineered E. coli strain in the study conveniently optimizes recombinant protein overexpression by suppressing cell autolysis, and shows potential industrial applications.

The composition dependence of the structural transition between the monoclinic 1T$^{\prime}$ and orthorhombic T$_{d}$ phases in the Mo$_{1-x}$W$_{x}$Te$_{2}$ Weyl semimetal was investigated by elastic neutron scattering on single crystals up to $x \approx 0.54$. First observed in MoTe$_{2}$, the transition from T$_{d}$ to 1T$^{\prime}$ is accompanied by an intermediate pseudo-orthorhombic phase, T$_{d}^{*}$. Upon doping with W, the T$_{d}^{*}$ phase vanishes by $x \approx 0.34$. Above this concentration, a phase coexistence behavior with both T$_{d}$ and 1T$^{\prime}$ is observed instead. The interlayer in-plane positioning parameter $\delta$, which relates to the 1T$^{\prime}$ $\beta$ angle, decreases with temperature as well as with W substitution, likely due to strong anharmonicity in the interlayer interactions. The temperature width of the phase coexistence remains almost constant up to $x \approx 0.54$, in contrast to the broadening reported under pressure., Comment: supplement in separate file

As a popular meta-learning approach, the model-agnostic meta-learning (MAML) algorithm has been widely used due to its simplicity and effectiveness. However, the convergence of the general multi-step MAML still remains unexplored. In this paper, we develop a new theoretical framework to provide such convergence guarantee for two types of objective functions that are of interest in practice: (a) resampling case (e.g., reinforcement learning), where loss functions take the form in expectation and new data are sampled as the algorithm runs; and (b) finite-sum case (e.g., supervised learning), where loss functions take the finite-sum form with given samples. For both cases, we characterize the convergence rate and the computational complexity to attain an $\epsilon$-accurate solution for multi-step MAML in the general nonconvex setting. In particular, our results suggest that an inner-stage stepsize needs to be chosen inversely proportional to the number $N$ of inner-stage steps in order for $N$-step MAML to have guaranteed convergence. From the technical perspective, we develop novel techniques to deal with the nested structure of the meta gradient for multi-step MAML, which can be of independent interest., Comment: 40 pages

For a long time asthma has been regarded as a lung disease associated with lung inflammation, airflow obstruction, coughing and mucus hypersecretion. Since their introduction 30 years ago, Inhalation corticosteroids and bronchodilators are still the cornerstone of asthma management nowadays, indicating that advancement in this field has somewhat stagnated. In recent years, we have started to appreciate that underneath the similar set of symptom manifestations, the asthma syndrome encompasses complex and heterogeneous pathophysiological mechanisms. The biologicals targeting specific type 2 cytokines have shown great clinical improvements in asthmatic patients and exemplify the importance of targeted therapy to advance asthma management. This thesis can be regarded as a continuous effort to identify undiscovered pathophysiological traits in asthma. We here investigated the therapeutic potential of C1-inhibitor in both human and murine asthma models, considering that C1-inhibitor exerts immunomodulatory effects as a regulator of both the complement and contact system. These systems have been implicated to play a role in the pathogenesis and pathophysiology of asthma, making them interesting for further research to elucidate their potential as candidates for targeted therapy.

The interfacial electronic structure of layer-by-layer 4-(dicyanomethylene)-2-t-butyl-6-(1,1,7,7-tetramethyljulolidyl-9-enyl)-4H-pyran (DCJTB)/fullerene C 60 was investigated using ultra-violet photoemission spectroscopy (UPS). Photoemission data of period of DCJTB/C 60 /DCJTB films suggested the formation of surface dipole and interfacial band bending across the interfaces, which greatly facilitates the charge transfer from DCJTB to C 60 and from C 60 to DCJTB layer as well. When applied this layer-by-layer structure to a near-infrared photodetector, a maximum of photocurrent was achieved by the device with 3 periods of DCJTB/C 60 thin films. Finally, the detailed work mechanism of this detector was discussed.

The rapid growth of the Internet of Things (IoT) is enhancing customers' experience with an ever-increasing mass of intelligent devices. However, the security doesn't get enough attention compared with novel concepts and features by manufactures which even not provide service of updates and patches. These issues could have been mitigated by the mechanism of the network access control. But traditional authentication mechanisms based on computational complexity of cryptographic protocols cannot be implemented on IoT devices with limited physical and computational resources. Therefore, we turn our attention to the device fingerprint which is like a biometric fingerprint that be able to provide identity authentication. The fingerprint proposed in this paper generated by network traffic features divided into the packet-feature and the IAT-feature. These features are extracted from the traffic when the device is booted, because at this time the device's network behavior is stable compared to other times. Experiments illustrate that the average identification accuracy of our device fingerprint reaches 92.4%

As mass devices access to network in the era of the Internet of Things (IoT), the network access control becomes more important. Traditional authentication mechanisms are no longer suitable for IoT devices with limited physical and computational resources because of their complexity. This problem can be solved by the device fingerprinting, which can identify devices only by their traffic characteristics. However, existing device fingerprinting technologies cannot accurately identify devices' type from the same manufacturer which equips these devices with similar hardwares and softwares. Therefore, we proposed the TSMC-SVM algorithm which introduced the cosine similarity into the SVM for improving the identification accuracy of these similar devices described above. Experiments illustrate that the average identification accuracy of the TSMC-SVM reaches 93.2%.

Clostridium beijerinckii is a potentially important industrial microorganism as it can synthesize valuable chemicals and fuels from various carbon sources. The establishment of convenient to use, effective gene tools with which the organism can be rapidly modified is essential if its full potential is to be realized. Here, we developed a genomic editing tool (pCBEclos) for use in C. beijerinckii based on the fusion of cytidine deaminase (Apobec1), Cas9 D10A nickase and uracil DNA glycosylase inhibitor (UGI). Apobec1 and UGI are guided to the target site where they introduce specific base‐pair substitutions through the conversion of C·G to T·A. By appropriate choice of target sequence, these nucleotide changes are capable of creating missense mutation or null mutations in a gene. Through optimization of pCBEclos, the system derived, pCBEclos‐opt, has been used to rapidly generate four different mutants in C. beijerinckii, in pyrE, xylR, spo0A, and araR. The efficiency of the system was such that they could sometimes be directly obtained following transformation, otherwise only requiring one single restreaking step. Whilst CRISPR–Cas9 nickase systems, such as pNICKclos2.0, have previously been reported in C. beijerinckii, pCBEclos‐opt does not rely on homologous recombination, a process that is intrinsically inefficient in clostridia such as C. beijerinckii. As a consequence, bulky editing templates do not need to be included in the knockout plasmids. This both reduces plasmid size and makes their construction simpler, for example, whereas the assembly of pNICKclos2.0 requires six primers for the assembly of a typical knockout plasmid, pCBEclos‐opt requires just two primers. The pCBEclos‐opt plasmid established here represents a powerful new tool for genome editing in C. beijerinckii, which should be readily applicable to other clostridial species.

Transformer-based pre-trained language models have proven to be effective for learning contextualized language representation. However, current approaches only take advantage of the output of the encoder's final layer when fine-tuning the downstream tasks. We argue that only taking single layer's output restricts the power of pre-trained representation. Thus we deepen the representation learned by the model by fusing the hidden representation in terms of an explicit HIdden Representation Extractor (HIRE), which automatically absorbs the complementary representation with respect to the output from the final layer. Utilizing RoBERTa as the backbone encoder, our proposed improvement over the pre-trained models is shown effective on multiple natural language understanding tasks and help our model rival with the state-of-the-art models on the GLUE benchmark.

Stochastic gradient descent (SGD) has been found to be surprisingly effective in training a variety of deep neural networks. However, there is still a lack of understanding on how and why SGD can train these complex networks towards a global minimum. In this study, we establish the convergence of SGD to a global minimum for nonconvex optimization problems that are commonly encountered in neural network training. Our argument exploits the following two important properties: 1) the training loss can achieve zero value (approximately), which has been widely observed in deep learning; 2) SGD follows a star-convex path, which is verified by various experiments in this paper. In such a context, our analysis shows that SGD, although has long been considered as a randomized algorithm, converges in an intrinsically deterministic manner to a global minimum., Comment: ICLR2019