Search

Abstract Statistical characterizations of complex network structures can be obtained from both the Ihara Zeta function (in terms of prime cycle frequencies) and the partition function from statistical mechanics. However, these two representations are usually regarded as separate tools for network analysis, without exploiting the potential synergies between them. In this paper, we establish a link between the Ihara Zeta function from algebraic graph theory and the partition function from statistical mechanics, and exploit this relationship to obtain a deeper structural characterisation of network structure. Specifically, the relationship allows us to explore the connection between the microscopic structure and the macroscopic characterisation of a network. We derive thermodynamic quantities describing the network, such as entropy, and show how these are related to the frequencies of prime cycles of various lengths. In particular, the n-th order partial derivative of the Ihara Zeta function can be used to compute the number of prime cycles in a network, which in turn is related to the partition function of Bose–Einstein statistics. The corresponding derived entropy allows us to explore a phase transition in the network structure with critical points at high and low-temperature limits. Numerical experiments and empirical data are presented to evaluate the qualitative and quantitative performance of the resulting structural network characterisations.

This study investigated the spatiotemporal distribution of the phytoplankton in the coral habitat of Dongshan Bay (China), along with critical factors affecting the distribution, during June, August, and December 2022. Phytoplankton abundance in Dongshan Bay exhibited considerably temporal variation, peaking in June 2022, gradually decreasing thereafter, and reaching its lowest point in December 2022. The abundance of bottom-layer phytoplankton consistently exceeded that of the surface layer throughout all seasons. The average phytoplankton abundance in the coral habitat of Dongshan Bay was lower than that in non-coral habitat areas. Fluctuations in the Zhangjiang River and coastal upwelling influenced the diversity and community structure of the phytoplankton. Critical factors causing spatiotemporal variability in phytoplankton community structure included nutrient concentrations and seawater temperature. Nutrients played key roles in influencing various phytoplankton groups. Dominant diatom species, such as Thalassionema nitzschioides and Thalassiosira diporocyclus, were positively correlated with ammonia nitrogen, seawater salinity, coral cover, and the number of coral species present. In winter, Calanus sinicus exhibited a negative correlation with harmful algal bloom species. Additionally, it was found that both in the coral habitat and surrounding open sea, currents, nutrients, and zooplankton may play crucial roles in determining the spatiotemporal variability in the phytoplankton community structure.

Climate change may threaten the survival of corals due to ocean warming and increased occurrence of extreme weather events. Yet, marginal coral communities demonstrate remarkable adaptability, largely due to stress-tolerant species. The mechanisms underlying their resilience to environmental fluctuations remains largely unexplored. Here, we investigated 3 coral species (Cyphastrea serailia, Dipsastraea speciosa, and Duncanopsammia peltata) living at the northern edge of their distribution along the South China coast (Dongshan). Combining physiological and stable isotope analysis, we assessed their traits (symbiont characteristics, C:N ratio, and biomass) and trophic flexibility (i.e., shifts in isotopic niche or nutritional mode) in response to environmental fluctuations across seasons (autumn, winter, and late spring). Our results revealed distinct physiological traits among these corals, but consistently showed lower symbiont density, higher biomass, and increased chlorophyll content in winter. All corals highly relied on heterotrophy as a crucial adaptive response (consistently no overlap between host and symbiont isotopic niches), but they differed in the degree of trophic plasticity, with D. speciosa showing the highest variability in heterotrophic contribution (36.3% across seasons), followed by C. serailia (21.8%) and D. peltata (8.6%). Notably, the species-specific differences in trophic plasticity well corresponded to the variation in overall isotopic niche and coral distribution. As such, trophic plasticity may play a role in shaping spatial distribution of corals in the marginal community. Overall, this study enhances our understanding of coral resilience in marginal communities and offer new insights into possible mechanisms driving the distribution of reef-building corals in the changing ocean.

Abstract The rapid development of social networking platforms has accelerated the spread of false information. Effective source location methods are essential to control the spread of false information. Most existing methods fail to make full use of the infection of neighborhood information in nodes, resulting in a poor source localization effect. In addition, most existing methods ignore the existence of multiple source nodes in the infected cluster and hard to identify the source nodes comprehensively. To solve these problems, we propose a new method about the multiple sources location with the neighborhood entropy. The method first defines the two kinds of entropy, i.e. infection adjacency entropy and infection intensity entropy, depending on whether neighbor nodes are infected or not. Then, the possibility of a node is evaluated by the neighborhood entropy. To locate the source nodes comprehensively, we propose a source location algorithm with the infected clusters. Other unrecognized source nodes in the infection cluster are identified by the cohesion of nodes, which can deal with the situation in the multiple source nodes in an infected cluster. We conduct experiments on various network topologies. Experimental results show that the two proposed algorithms outperform the existing methods.

Global change has generated challenges for oceans, from individuals to the entire ecosystem, and has raised contemporary issues related to ocean conservation and management. Specifically, coral reef ecosystems have been exposed to various environmental and human disturbances. In this study, the Ecopath with Ecosim model was used to explore the impacts of ocean warming and fishing on Xisha Islands coral reef ecosystem in the South China Sea. The variables in this model included two ocean warming scenarios and three fishing scenarios. The model consisted of 23 functional groups including algae, coral, sea birds, and sharks. Our results showed that by the middle of the century, ocean warming and fishing led to a 3.79% and 4.74% decrease in total catch compared with 2009, respectively. In addition, the combined effects of ocean warming and fishing caused a 4.79% decrease in total catch, and the mean trophic level of catch was predicted to decrease by 6.01% under the SSP585-High fishing scenario. Reducing the fishing effort mitigates the effects of ocean warming on some species, such as large carnivorous fish and medium carnivorous fish; however, under low fishing effort, some functional groups, such as small carnivorous and omnivorous fish, have low biomass because of higher predation mortality.

The species Endeis straughani Clark, 1970 had only been reported from Queensland (Australia) until specimens were collected in Ghana (West Africa) which were initially described as Endeis picta Bamber, 1979.Five immature individuals of E. straughani were gathered amongst hydroids during the 5th leg of the DY125-34 expedition. It was speculated that these were from Port Louis (Mauritius) and were a new record for this rarely-collected species.

Complete mitochondrial DNA sequence data have played a significant role in phylogenetic and evolutionary studies of scleractinian corals. In this study, the complete mitogenome of Psammocora profundacella Gardiner, 1898, collected from Guangdong Province, China, was sequenced by next-generation sequencing for the first time. Psammocora profundacella is the first species for which a mitogenome has been sequenced in the family Psammocoridae. The length of its assembled mitogenome sequence was 16,274 bp, including 13 protein-coding genes, two tRNAs and two rRNAs. Its gene content and gene order were consistent with the other Scleractinia species. All genes were encoded on the H strand and the GC content of the mitochondrial genome was 30.49%. Gene content and order were consistent with the other Scleractinia species. Based on 13 protein-coding genes, Maximum Likelihood phylogenetic analysis showed that P. profundacella belongs to the “Robust” clade. Mitochondrial genome data provide important molecular information for understanding the phylogeny of stony corals. More variable markers and additional species should be sequenced to confirm the evolutionary relationships of Scleractinia in the future.

In this study, the complete mitogenome sequence of the stony coral, Turbinaria bifrons Brüggemann, 1877, has been decoded for the first time by next-generation sequencing (NGS) and genome assembly. The assembled mitogenome was 18,880 bp in length, contained 13 protein coding genes, 2 transfer RNAs, and 2 ribosomal RNAs. The complete mitogenome of T. bifrons showing 97.09% identities to Tubastraea tagusensis. The complete mitogenome provides essential and important DNA molecular data for further phylogenetic and evolutionary analysis for coral phylogeny.

Large-scale automatic speech recognition model has achieved impressive performance. However, huge computational resources and massive amount of data are required to train an ASR model. Knowledge distillation is a prevalent model compression method which transfers the knowledge from large model to small model. To improve the efficiency of knowledge distillation for end-to-end speech recognition especially in the low-resource setting, a Mixup-based Knowledge Distillation (MKD) method is proposed which combines Mixup, a data-agnostic data augmentation method, with softmax-level knowledge distillation. A loss-level mixture is presented to address the problem caused by the non-linearity of label in the KL-divergence when adopting Mixup to the teacher–student framework. It is mathematically shown that optimizing the mixture of loss function is equivalent to optimize an upper bound of the original knowledge distillation loss. The proposed MKD takes the advantage of Mixup and brings robustness to the model even with a small amount of training data. The experiments on Aishell-1 show that MKD obtains a 15.6% and 3.3% relative improvement on two student models with different parameter scales compared with the existing methods. Experiments on data efficiency demonstrate MKD achieves similar results with only half of the original dataset.

Over the past few decades, modern coral taxonomy, combining morphology and molecular sequence data, has resolved many long-standing questions about scleractinian corals. In this study, we sequenced the complete mitochondrial genomes of three Merulinidae corals (Dipsastraea rotumana, Favites pentagona, and Hydnophora exesa) for the first time using next-generation sequencing. The obtained mitogenome sequences ranged from 16,466 bp (D. rotumana) to 18,006 bp (F. pentagona) in length, and included 13 unique protein-coding genes (PCGs), two transfer RNA genes, and two ribosomal RNA genes . Gene arrangement, nucleotide composition, and nucleotide bias of the three Merulinidae corals were canonically identical to each other and consistent with other scleractinian corals. We performed a Bayesian phylogenetic reconstruction based on 13 protein-coding sequences of 86 Scleractinia species. The results showed that the family Merulinidae was conventionally nested within the robust branch, with H. exesa clustered closely with F. pentagona and D. rotumana clustered closely with Favites abdita. This study provides novel insight into the phylogenetics of species within the family Merulinidae and the evolutionary relationships among different Scleractinia genera.

We sincerely apologize for the inconvenience of updating the authorship [...]

The stock market is a dynamical system composed of intricate relationships between financial entities, such as banks, corporations, and institutions. Such a complex interactive system can be represented by the network structure. The underlying mechanism of stock exchange establishes a time-evolving network among companies and individuals, which characterise the correlations of stock prices in the time sequential trades. Here, we develop a novel technique in quantum statistics to analyse the financial market evolution. We commence from heat bath analogy where the normalised Laplacian matrix plays the role of the Hamiltonian operator of the network. The eigenvalues of the Hamiltonian specify energy states of the network. These states are occupied by either indistinguishable bosons or fermions with corresponding Bose-Einstein and Fermi-Dirac statistics. Using the relevant partition functions, we develop the thermodynamic entropy to explore dynamic network characterisations. We conduct the experiments to apply this novel method to identify the significant variance in network structure during the financial crisis. The thermodynamic entropy provides an excellent framework to represent the variations taking place in the stock market.

This paper seeks to advance the state-of-the-art in analysing fMRI data to detect onset of Alzheimer’s disease and identify stages in the disease progression. We employ methods of network neuroscience to represent correlation across fMRI data arrays, and introduce novel techniques for network construction and analysis. In network construction, we vary thresholds in establishing BOLD time series correlation between nodes, yielding variations in topological and other network characteristics. For network analysis, we employ methods developed for modelling statistical ensembles of virtual particles in thermal systems. The microcanonical ensemble and the canonical ensemble are analogous to two different fMRI network representations. In the former case, there is zero variance in the number of edges in each network, while in the latter case the set of networks have a variance in the number of edges. Ensemble methods describe the macroscopic properties of a network by considering the underlying microscopic characterisations which are in turn closely related to the degree configuration and network entropy. When applied to fMRI data in populations of Alzheimer’s patients and controls, our methods demonstrated levels of sensitivity adequate for clinical purposes in both identifying brain regions undergoing pathological changes and in revealing the dynamics of such changes.

Federated learning (FL) is a privacy-preserving technique for training a vast amount of decentralized data and making inferences on mobile devices. As a typical language modeling problem, mobile keyboard prediction aims at suggesting a probable next word or phrase and facilitating the human-machine interaction in a virtual keyboard of the smartphone or laptop. Mobile keyboard prediction with FL hopes to satisfy the growing demand that high-level data privacy be preserved in artificial intelligence applications even with the distributed models training. However, there are two major problems in the federated optimization for the prediction: (1) aggregating model parameters on the server-side and (2) reducing communication costs caused by model weights collection. To address the above issues, traditional FL methods simply use averaging aggregation or ignore communication costs. We propose a novel Federated Mediation (FedMed) framework with the adaptive aggregation, mediation incentive scheme, and topK strategy to address the model aggregation and communication costs. The performance is evaluated in terms of perplexity and communication rounds. Experiments are conducted on three datasets (i.e., Penn Treebank, WikiText-2, and Yelp) and the results demonstrate that our FedMed framework achieves robust performance and outperforms baseline approaches.

With the advent of the Internet of things (IoT), intelligent transportation has evolved over time to improve traffic safety and efficiency as well as to reduce congestion and environmental pollution. However, there are some challenging issues to be addressed so that it can be implemented to its full potential. The major challenge in intelligent transportation is that vehicles and pedestrians, as the main types of edge nodes in IoT infrastructure, are on the constant move. Hence, the topology of the large scale network is changing rapidly over time and the service chain may need reestablishment frequently. Existing Virtual Network Function (VNF) chain placement methods are mostly good at static network topology and any evolvement of the network requires global computation, which leads to the inefficiency in computing and the waste of resources. Mapping the network topology to a graph, we propose a novel VNF placement method called BVCP (Border VNF Chain Placement) to address this problem by elaborately dividing the graph into multiple subgraphs and fully exploiting border hypervisors. Experimental results show that BVCP outperforms the state-of-the-art method in VNF chain placement, which is highly efficient in large scale IoT of intelligent transportation.

Lack of mitochondrial genome data of Scleractinia is hampering progress across genetic, systematic, phylogenetic and evolutionary studies. In this study, the complete mitogenome sequence of the stony coral, Acropora valida (Dana, 1846), has been decoded for the first time by next-generation sequencing and genome assembly. The assembled mitogenome was 18,385 bp in length, containing 13 protein-coding genes (PCGs), two transfer RNAs and two ribosomal RNAs. All genes were encoded on the same strand. The A + T content of the mitochondrial genome was 62.0% (25.0% A, 37.0% T, 24.3% G, and 13.8% for C). The complete mitogenome of Acropora valida has the same gene content and structure as other Acropora species. The complete mitogenome provides essential and important DNA molecular data for further phylogenetic and evolutionary analyses for coral phylogeny.

In this study, the complete mitogenome sequence of Pentacheles validus has been decoded for the first time. The overall base composition is 33.4% for A, 25.0% for C, 10.8% for G, and 30.9% for T and has low GC content of 35.8%. The assembled mitogenome, 16,079 bp in length, has unique 13 protein-coding genes (PCGs), 22 transfer RNAs and 2 ribosomal RNAs. The mitogenome shares 80% identity to Polycheles typhlops. The complete mitogenome of P. validus provides essential DNA molecular data for further phylogenetic and evolutionary analysis for Polychelidae and deep-sea faunas.

A new species of pycnogonid collected by the Chinese research vessel R/V Dayangyihao during cruises to the Southwest Indian Ridge in 2008 and 2009 is recorded. The new species, Austrodecus bamberi, is placed into the tristanense-section by the characters of 4-articled ovigers and present auxiliary claws and is distinguished from other species in this section by the number and length of tubercles on the first coxae.

The particular environmental characteristics of deep water such as its immense scale and high pressure systems, presents technological problems that have prevented research to broaden our knowledge of deep-sea fish. Here, we described the mitogenome sequence of a deep-sea fish, Cetonurus globiceps. The genome is 17,137 bp in length, with a standard set of 22 transfer RNA genes (tRNAs), two ribosomal RNA genes, 13 protein-coding genes, and two typical non-coding control regions. Additionally, a 70 bp tRNA(Thr)-tRNA(Pro) intergenic spacer is present. The C. globiceps mitogenome exhibited strand-specific asymmetry in nucleotide composition. The AT-skew and GC-skew values in the whole genome of C. globiceps were 0 and -0.2877, respectively, revealing that the H-strand had equal amounts of A and T and that the overall nucleotide composition was C skewed. All of the tRNA genes could be folded into cloverleaf secondary structures, while the secondary structure of tRNA(Ser(AGY)) lacked a discernible dihydrouridine stem. By comparing this genome sequence with the recognition sites in teleost species, several conserved sequence blocks were identified in the control region. However, the GTGGG-box, the typical characteristic of conserved sequence block E (CSB-E), was absent. Notably, tandem repeats were identified in the 3' portion of the control region. No similar repetitive motifs are present in most of other gadiform species. Phylogenetic analysis based on 12 protein coding genes provided strong support that C. globiceps was the most derived in the clade. Some relationships however, are in contrast with those presented in previous studies. This study enriches our knowledge of mitogenomes of the genus Cetonurus and provides valuable information on the evolution of Macrouridae mtDNA and deep-sea fish.