There is a growing trend to design hybrid neural networks (HNNs) by combining spiking neural networks and artificial neural networks to leverage the strengths of both. Here, we propose a framework ...for general design and computation of HNNs by introducing hybrid units (HUs) as a linkage interface. The framework not only integrates key features of these computing paradigms but also decouples them to improve flexibility and efficiency. HUs are designable and learnable to promote transmission and modulation of hybrid information flows in HNNs. Through three cases, we demonstrate that the framework can facilitate hybrid model design. The hybrid sensing network implements multi-pathway sensing, achieving high tracking accuracy and energy efficiency. The hybrid modulation network implements hierarchical information abstraction, enabling meta-continual learning of multiple tasks. The hybrid reasoning network performs multimodal reasoning in an interpretable, robust and parallel manner. This study advances cross-paradigm modeling for a broad range of intelligent tasks.
Epidemiologic and medical studies often rely on evaluators to obtain measurements of exposures or outcomes for study participants, and valid estimates of associations depends on the quality of data. ...Even though statistical methods have been proposed to adjust for measurement errors, they often rely on unverifiable assumptions and could lead to biased estimates if those assumptions are violated. Therefore, methods for detecting potential 'outlier' evaluators are needed to improve data quality during data collection stage.
In this paper, we propose a two-stage algorithm to detect 'outlier' evaluators whose evaluation results tend to be higher or lower than their counterparts. In the first stage, evaluators' effects are obtained by fitting a regression model. In the second stage, hypothesis tests are performed to detect 'outlier' evaluators, where we consider both the power of each hypothesis test and the false discovery rate (FDR) among all tests. We conduct an extensive simulation study to evaluate the proposed method, and illustrate the method by detecting potential 'outlier' audiologists in the data collection stage for the Audiology Assessment Arm of the Conservation of Hearing Study, an epidemiologic study for examining risk factors of hearing loss in the Nurses' Health Study II.
Our simulation study shows that our method not only can detect true 'outlier' evaluators, but also is less likely to falsely reject true 'normal' evaluators.
Our two-stage 'outlier' detection algorithm is a flexible approach that can effectively detect 'outlier' evaluators, and thus data quality can be improved during data collection stage.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
Accurate identification of Seriola dumerili (SD) offers crucial technical support for aquaculture practices and behavioral research of this species. However, the task of discerning S. dumerili from ...complex underwater settings, fluctuating light conditions, and schools of fish presents a challenge. This paper proposes an intelligent recognition model based on the YOLOv8 network called SD-YOLOv8. By adding a small object detection layer and head, our model has a positive impact on the recognition capabilities for both close and distant instances of S. dumerili, significantly improving them. We construct a convenient S. dumerili dataset and introduce the deformable convolution network v2 (DCNv2) to enhance the information extraction process. Additionally, we employ the bottleneck attention module (BAM) and redesign the spatial pyramid pooling fusion (SPPF) for multidimensional feature extraction and fusion. The Inner-MPDIoU bounding box regression function adjusts the scale factor and evaluates geometric ratios to improve box positioning accuracy. The experimental results show that our SD-YOLOv8 model achieves higher accuracy and average precision, increasing from 89.2% to 93.2% and from 92.2% to 95.7%, respectively. Overall, our model enhances detection accuracy, providing a reliable foundation for the accurate detection of fishes.
There are two principle approaches for learning in artificial intelligence: error-driven global learning and neuroscience-oriented local learning. Integrating them into one network may provide ...complementary learning capabilities for versatile learning scenarios. At the same time, neuromorphic computing holds great promise, but still needs plenty of useful algorithms and algorithm-hardware co-designs to fully exploit its advantages. Here, we present a neuromorphic global-local synergic learning model by introducing a brain-inspired meta-learning paradigm and a differentiable spiking model incorporating neuronal dynamics and synaptic plasticity. It can meta-learn local plasticity and receive top-down supervision information for multiscale learning. We demonstrate the advantages of this model in multiple different tasks, including few-shot learning, continual learning, and fault-tolerance learning in neuromorphic vision sensors. It achieves significantly higher performance than single-learning methods. We further implement the model in the Tianjic neuromorphic platform by exploiting algorithm-hardware co-designs and prove that the model can fully utilize neuromorphic many-core architecture to develop hybrid computation paradigm.
Elevated excitability of glutamatergic neurons in the lateral parabrachial nucleus (PBL) is associated with the pathogenesis of inflammatory pain, but the underlying molecular mechanisms are not ...fully understood. Sodium leak channel (NALCN) is widely expressed in the central nervous system and regulates neuronal excitability. In this study, chemogenetic manipulation was used to explore the association between the activity of PBL glutamatergic neurons and pain thresholds. Complete Freund's adjuvant (CFA) was used to construct an inflammatory pain model in mice. Pain behaviour was tested using von Frey filaments and Hargreaves tests. Local field potential (LFP) was used to record the activity of PBL glutamatergic neurons. Gene knockdown techniques were used to investigate the role of NALCN in inflammatory pain. We further explored the downstream projections of PBL using cis-trans-synaptic tracer virus. The results showed that chemogenetic inhibition of PBL glutamatergic neurons increased pain thresholds in mice, whereas chemogenetic activation produced the opposite results. CFA plantar modelling increased the number of C-Fos protein and NALCN expression in PBL glutamatergic neurons. Knockdown of NALCN in PBL glutamatergic neurons alleviated CFA-induced pain. CFA injection induced C-Fos protein expression in central nucleus amygdala (CeA) neurons, which was suppressed by NALCN knockdown in PBL glutamatergic neurons. Therefore, elevated expression of NALCN in PBL glutamatergic neurons contributes to the development of inflammatory pain via PBL-CeA projections.
Soil cadmium (Cd) contamination is a serious problem on agricultural land. Adequate nitrogen (N) may help ameliorate plant fitness under Cd stress. This study examined the role of N application in ...improving maize tolerance to Cd stress. Two maize genotypes, Zhongke11 (larger root system) and Shengrui999 (smaller root system), were grown in a loessal soil amended with Cd (Cd0, no added Cd; Cd1, 20 mg kg−1 soil as CdCl2·2.5 H2O) and N (N0, no added N; N1, 100 mg kg−1 soil as urea) under greenhouse, and plants were assessed at silking and maturity stages. Maize plants exhibited moderate Cd stress with significantly reduced grain yield, especially under low N (N1). Roots accumulated more Cd than above-ground parts. Grain Cd concentration was the least (0.05–0.06 μg g−1) among all organs which is below the safety threshold. Leaf Cd concentrations (0.24–1.18 mg kg−1) were also under the toxicity threshold. Nitrogen addition significantly improved plant growth, chlorophyll content, photosynthesis traits, and tissue Cd contents, and reduced Cd concentration in soil compared to N0 treatment. Nitrogen promoted Cd bioconcentration and translocation factors in stem and leaves. Cadmium stress reduced N fertilizer agronomic efficiency at maturity. At maturity, root Cd content was positively correlated with root N and calcium accumulation, and stem Cd content was positively correlated with stem N content (both P ≤ 0.05). Genotypes with different root system size differed in response to Cd toxicity and / or N deficit. The small-rooted genotype Shengrui999 was more tolerant to moderate Cd stress than the large-rooted Zhongke11. Addition of N ameliorated Cd stress in both maize genotypes by improving plant growth performance, and regulating Cd translocations among plant organs.
Display omitted
•N supply improved plant growth and Cd tolerance by regulating Cd translocations.•Small-rooted genotype Shengrui999 was more tolerant to moderate Cd stress than large-rooted Zhongke11.•Cd stress reduced N fertilizer agronomic efficiency.•Interactive effects of low N and Cd stress reduced plant growth and yield.
It is widely believed the brain-inspired spiking neural networks have the capability of processing temporal information owing to their dynamic attributes. However, how to understand what kind of ...mechanisms contributing to the learning ability and exploit the rich dynamic properties of spiking neural networks to satisfactorily solve complex temporal computing tasks in practice still remains to be explored. In this article, we identify the importance of capturing the multi-timescale components, based on which a multi-compartment spiking neural model with temporal dendritic heterogeneity, is proposed. The model enables multi-timescale dynamics by automatically learning heterogeneous timing factors on different dendritic branches. Two breakthroughs are made through extensive experiments: the working mechanism of the proposed model is revealed via an elaborated temporal spiking XOR problem to analyze the temporal feature integration at different levels; comprehensive performance benefits of the model over ordinary spiking neural networks are achieved on several temporal computing benchmarks for speech recognition, visual recognition, electroencephalogram signal recognition, and robot place recognition, which shows the best-reported accuracy and model compactness, promising robustness and generalization, and high execution efficiency on neuromorphic hardware. This work moves neuromorphic computing a significant step toward real-world applications by appropriately exploiting biological observations.
The accurate prediction of protein–ligand binding free energies remains a significant challenge of central importance in computational biophysics and structure-based drug design. Multiple recent ...advances including the development of greatly improved protein and ligand molecular mechanics force fields, more efficient enhanced sampling methods, and low-cost powerful GPU computing clusters have enabled accurate and reliable predictions of relative protein–ligand binding free energies through the free energy perturbation (FEP) methods. However, the existing FEP methods can only be used to calculate the relative binding free energies for R-group modifications or single-atom modifications and cannot be used to efficiently evaluate scaffold hopping modifications to a lead molecule. Scaffold hopping or core hopping, a very common design strategy in drug discovery projects, is critical not only in the early stages of a discovery campaign where novel active matter must be identified but also in lead optimization where the resolution of a variety of ADME/Tox problems may require identification of a novel core structure. In this paper, we introduce a method that enables theoretically rigorous, yet computationally tractable, relative protein–ligand binding free energy calculations to be pursued for scaffold hopping modifications. We apply the method to six pharmaceutically interesting cases where diverse types of scaffold hopping modifications were required to identify the drug molecules ultimately sent into the clinic. For these six diverse cases, the predicted binding affinities were in close agreement with experiment, demonstrating the wide applicability and the significant impact Core Hopping FEP may provide in drug discovery projects.
An environmentally sustainable method to extract phosphatidylcholine (PC) from chicken liver (PCCL) and its functional properties were studied. The extraction times, enzymatic hydrolysis time, the ...solid-liquid ratio as well as types of enzymes (protamex proteinase and neutral proteinase) were investigated. Furthermore, the content of PCCL, emulsifying properties and solubilities of PCCL were also determined. The optimum conditions of extracting PCCL were found to be: reaction time of 3.75 h, enzymatic hydrolysis time of 85.22 min, 1: 3.15 (w/v) of solid-liquid ratio, using protamex proteinase, and the yield and concentration of PCCL was 88.92% and 0.89 mg/mL, respectively. Solubility and emulsifying properties of PCCL showed that the HLB value of PCCL was 10, and in ethanol and glycerol, the solubility of PCCL was 0.5850 g/mL and 0.0965 g/mL, respectively, which was shown to have good ethanol solubility and lipophilicity. From the perspective of green production and high-value utilization of by-products, PCCL could be used as a potential new lecithin source, providing ideas for the development and application of PC of animal origin.
Abstract
Background
Childbearing in women with advanced maternal age (AMA) has increased the need for artificial reproductive technology (ART). ART and oxidative stress are associated with many ...pregnancy complications. Paraoxonase (PON) 1 is one of the key components responsible for antioxidative activity in high-density lipoprotein (HDL). This study aimed to investigate the longitudinal changes of oxidative stress and PON1 lactonase activity and status in older women undergoing ART.
Methods
This prospective nested case-control study included 129 control and 64 ART women. Blood samples were obtained respectively at different stages of pregnancy. PON1 level and lactonase activity were assessed using 7-O-diethylphosphoryl-3-cyano-4-methyl-7-hydroxycoumarin (DEPCyMC) and 5-thiobutyl butyrolactone (TBBL) as a substrate, respectively. A normalized lactonase activity (NLA) was estimated based on the ratio of TBBLase to DEPCyMCase activity. Serum total oxidant status (TOS), total antioxidant capacity (TAC), malondialdehyde (MDA), homocysteine (HCY),
PON1
C-108T and Q192R genetic polymorphisms, and metabolic parameters were analyzed.
Results
Lactonase activity and level of PON1 gradually decreased with pregnancy progression, while glycolipid metabolism parameters and TAC levels increased with pregnancy progression or significantly raised during the 2nd and 3rd trimesters, and NLA of PON1, TOS, OSI, MDA, and HCY significantly increased before delivery in the ART and control groups. Compared with the control women, the ART women had substantially higher or relatively high lactonase activity and NLA of PON1 and TAC during pregnancy; higher triglyceride (TG), total cholesterol, low-density lipoprotein cholesterol, atherogenic index, apolipoprotein (apo) B, and apoB/apoA1 ratio in the 1st trimester; and higher fasting glucose, fasting insulin, homeostatic model assessment of insulin resistance, and TG levels before delivery. No significant differences were found in the frequencies of
PON1
C-108T and Q192R genotypes and alleles between the ART and control groups.
Conclusions
Women with AMA undergoing ART had higher TAC, PON1 lactonase activity, and PON1 NLA than control women, suggesting increased compensatory antioxidant capacity in ART women, thus showing higher sensitivity to oxidative stress-related injury and diseases.
Celotno besedilo
Dostopno za:
DOBA, IZUM, KILJ, NUK, PILJ, PNG, SAZU, SIK, UILJ, UKNU, UL, UM, UPUK
PDF
