Community detection algorithms: a comparative analysis

The scholar's best friend: research trends in dog cognitive and behavioral studies

In recent decades, cognitive and behavioral knowledge in dogs seems to have developed considerably, as deduced from the published peer-reviewed articles. However, to date, the worldwide trend of scientific research on dog cognition and behavior has never been explored using a bibliometric approach, while the evaluation of scientific research has increasingly become important in recent years. In this review, we compared the publication trend of the articles in the last 34 years on dogs’ cognitive and behavioral science with those in the general category “Behavioral Science”. We found that, after 2005, there has been a sharp increase in scientific publications on dogs. Therefore, the year 2005 has been used as “starting point” to perform an in-depth bibliometric analysis of the scientific activity in dog cognitive and behavioral studies. The period between 2006 and 2018 is taken as the study period, and a backward analysis was also carried out. The data analysis was performed using “bibliometrix”, a new R-tool used for comprehensive science mapping analysis. We analyzed all information related to sources, countries, affiliations, co-occurrence network, thematic maps, collaboration network, and world map. The results scientifically support the common perception that dogs are attracting the interest of scholars much more now than before and more than the general trend in cognitive and behavioral studies. Both, the changes in research themes and new research themes, contributed to the increase in the scientific production on the cognitive and behavioral aspects of dogs. Our investigation may benefit the researchers interested in the field of cognitive and behavioral science in dogs, thus favoring future research work and promoting interdisciplinary collaborations.

Fuzzy Measures: A solution to deal with community detection problems for networks with additional information

In this work we introduce the notion of the weighted graph associated with a fuzzy measure. Having a finite set of elements between which there exists an affinity fuzzy relation, we propose the definition of a group based on that affinity fuzzy relation between the individuals. Then, we propose an algorithm based on the Louvain’s method to deal with community detection problems with additional information independent of the graph. We also provide a particular method to solve community detection problems over extended fuzzy graphs. Finally, we test the performance of our proposal by means of some detailed computational tests calculated in several benchmark models.

Spatiotemporal Exploration of Chinese Spring Festival Population Flow Patterns and Their Determinants Based on Spatial Interaction Model

Large-scale population flow reshapes the economic landscape and is affected by unbalanced urban development. The exploration of migration patterns and their determinants is therefore crucial to reveal unbalanced urban development. However, low-resolution migration datasets and insufficient consideration of interactive differences have limited such exploration. Accordingly, based on 2019 Chinese Spring Festival travel-related big data from the AMAP platform, we used social network analysis (SNA) methods to accurately reveal population flow patterns. Then, with consideration of the spatial heterogeneity of interactive patterns, we used spatially weighted interactive models (SWIMs), which were improved by the incorporation of weightings into the global Poisson gravity model, to efficiently quantify the effect of socioeconomic factors on migration patterns. These SWIMs generated the local characteristics of the interactions and quantified results that were more regionally consistent than those generated by other spatial interaction models. The migration patterns had a spatially vertical structure, with the city development level being highly consistent with the flow intensity; for example, the first-level developments of Beijing, Shanghai, Chengdu, Guangzhou, Shenzhen, and Chongqing occupied a core position. A spatially horizontal structure was also formed, comprising 16 closely related city communities. Moreover, the quantified impact results indicated that migration pattern variation was significantly related to the population, value-added primary and secondary industry, the average wage, foreign capital, pension insurance, and certain aspects of unbalanced urban development. These findings can help policymakers to guide population migration, rationally allocate industrial infrastructure, and balance urban development.

The sustainability and the survivability of Kyoto's traditional craft industry revealed from supplier-customer network

Due to the changes in consumer demand and generational transformations, Kyoto’s traditional craft industry has suffered substantial revenue losses in recent years. This research aimed to characterize Kyoto’s traditional craft industry by analyzing the supplier-customer network involving individual firms within the Kyoto region. In the process, we clarify the community structure, key firms, network topological characteristics, bow-tie structure, robustness, the vulnerability of the supplier-customer network as crucial factors for sustainable growth. The community and bow-tie structure analysis became clear that the traditional craft industry continues to occupy an important position in Kyoto’s industrial network. Furthermore, we clarify the relationship between modern and traditional craft industries’ network characteristics and their relative profitability and productivity. It became evident that the traditional craft industry has a different network structure from the modern consumer games and electric machinery industries. The modern industries have the strongly coupled component, and the attendant firms there create high value-added and play a significant role in driving the entire industry, while more traditional craft industries, such as the Nishijin silk fabrics and Kyoto doll industries, do not have this strongly coupled component. Moreover, the traditional crafts industry does not have a central firm or a dense network for integrating information, which is presumed to be a factor in the decline of the traditional craft industry.

The network structure and eco-evolutionary dynamics of CRISPR-induced immune diversification

As a heritable sequence-specific adaptive immune system, CRISPR-Cas is a powerful force shaping strain diversity in host-virus systems. While the diversity of CRISPR alleles has been explored, the associated structure and dynamics of host-virus interactions have not. We explore the role of CRISPR in mediating the interplay between host-virus interaction structure and eco-evolutionary dynamics in a computational model and compare the results with three empirical datasets from natural systems. We show that the structure of the networks describing who infects whom and the degree to which strains are immune, are respectively modular (containing groups of hosts and viruses that interact strongly) and weighted-nested (specialist hosts are more susceptible to subsets of viruses that in turn also infect the more generalist hosts with many spacers matching many viruses). The dynamic interplay between these networks influences transitions between dynamical regimes of virus diversification and host control. The three empirical systems exhibit weighted-nested immunity networks, a pattern our theory shows is indicative of hosts able to suppress virus diversification. Previously missing from studies of microbial host-pathogen systems, the immunity network plays a key role in the coevolutionary dynamics.

Detecting brain network communities: Considering the role of information flow and its different temporal scales

The identification of community structure in graphs continues to attract great interest in several fields. Network neuroscience is particularly concerned with this problem considering the key roles communities play in brain processes and functionality. Most methods used for community detection in brain graphs are based on the maximization of a parameter-dependent modularity function that often obscures the physical meaning and hierarchical organization of the partitions of network nodes. In this work, we present a new method able to detect communities at different scales in a natural, unrestricted way. First, to obtain an estimation of the information flow in the network we release random walkers to freely move over it. The activity of the walkers is separated into oscillatory modes by using empirical mode decomposition. After grouping nodes by their co-occurrence at each time scale, k-modes clustering returns the desired partitions. Our algorithm was first tested on benchmark graphs with favorable performance. Next, it was applied to real and simulated anatomical and/or functional connectomes in the macaque and human brains. We found a clear hierarchical repertoire of community structures in both the anatomical and the functional networks. The observed partitions range from the evident division in two hemispheres -in which all processes are managed globally- to specialized communities seemingly shaped by physical proximity and shared function. Additionally, the spatial scales of a network's community structure (characterized by a measure we term within-communities path length) appear inversely proportional to the oscillatory modes' average frequencies. The proportionality constant may constitute a network-specific propagation velocity for the information flow. Our results stimulate the research of hierarchical community organization in terms of temporal scales of information flow in the brain network.

Rare events and first passage time statistics from the energy landscape

We analyze the probability distribution of rare first passage times corresponding to transitions between product and reactant states in a kinetic transition network. The mean first passage times and the corresponding rate constants are analyzed in detail for two model landscapes and the double funnel landscape corresponding to an atomic cluster. Evaluation schemes based on eigendecomposition and kinetic path sampling, which both allow access to the first passage time distribution, are benchmarked against mean first passage times calculated using graph transformation. Numerical precision issues severely limit the useful temperature range for eigendecomposition, but kinetic path sampling is capable of extending the first passage time analysis to lower temperatures, where the kinetics of interest constitute rare events. We then investigate the influence of free energy based state regrouping schemes for the underlying network. Alternative formulations of the effective transition rates for a given regrouping are compared in detail to determine their numerical stability and capability to reproduce the true kinetics, including recent coarse-graining approaches that preserve occupancy cross correlation functions. We find that appropriate regrouping of states under the simplest local equilibrium approximation can provide reduced transition networks with useful accuracy at somewhat lower temperatures. Finally, a method is provided to systematically interpolate between the local equilibrium approximation and exact intergroup dynamics. Spectral analysis is applied to each grouping of states, employing a moment-based mode selection criterion to produce a reduced state space, which does not require any spectral gap to exist, but reduces to gap-based coarse graining as a special case. Implementations of the developed methods are freely available online.

Testing "efficient supply chain propositions" using topological characterization of the global supply chain network

In this paper, we study the topological properties of the global supply chain network in terms of its degree distribution, clustering coefficient, degree-degree correlation, bow-tie structure, and community structure to test the efficient supply chain propositions proposed by E. J.S. Hearnshaw et al. The global supply chain data in the year 2017 are constructed by collecting various company data from the web site of Standard & Poor’s Capital IQ platform. The in- and out-degree distributions are characterized by a power law of the form of γ_in = 2.42 and γ_out = 2.11. The clustering coefficient decays 〈C(k)〉∼k-βk with an exponent β_k = 0.46. The nodal degree-degree correlations 〈k_nn(k)〉 indicates the absence of assortativity. The bow-tie structure of giant weakly connected component (GWCC) reveals that the OUT component is the largest and consists 41.1% of all firms. The giant strong connected component (GSCC) is comprised of 16.4% of all firms. We observe that upstream or downstream firms are located a few steps away from the GSCC. Furthermore, we uncover the community structures of the network and characterize them according to their location and industry classification. We observe that the largest community consists of the consumer discretionary sector based mainly in the United States (US). These firms belong to the OUT component in the bow-tie structure of the global supply chain network. Finally, we confirm the validity of Hearnshaw et al.’s efficient supply chain propositions, namely Proposition S1 (short path length), Proposition S2 (power-law degree distribution), Proposition S3 (high clustering coefficient), Proposition S4 (“fit-gets-richer” growth mechanism), Proposition S5 (truncation of power-law degree distribution), and Proposition S7 (community structure with overlapping boundaries) regarding the global supply chain network. While the original propositions S1 just mentioned a short path length, we found the short path from the GSCC to IN and OUT by analyzing the bow-tie structure. Therefore, the short path length in the bow-tie structure is a conceptual addition to the original propositions of Hearnshaw.

Network representations of diversity in scientific teams

One strength of network analysis is its ability to encapsulate social heterogeneity. Here, we leverage that strength to examine another dimension of individual heterogeneity: heterogeneity of skills, knowledge and experience. This skill heterogeneity is difficult to quantify, but is vitally important to outcomes for both individuals and teams. Complicating the matter, skill diversity can be present on multiple levels. Individuals have different kinds of skills, but they also have different degrees of specialization. Skill diversity on a team level may come from individual skill diversity or focused researchers in different areas. Here, we illustrate our network-based method for characterizing skill sets in a context of increasing importance: scientific collaboration. Using data from the field of economics, we create network-based measures of paper scope, individual specialization, coauthor alignment and team skill diversity. We then use those measures to examine the relationship between skill diversity and publication outcomes.

A Guide for Choosing Community Detection Algorithms in Social Network Studies: The Question Alignment Approach

INTRODUCTION

Community detection, the process of identifying subgroups of highly connected individuals within a network, is an aspect of social network analysis that is relevant but potentially underutilized in prevention research. Guidance on using community detection methods stresses aligning methods with specific research questions but lacks clear operationalization. The Question Alignment approach was developed to help address this gap and promote the high-quality use of community detection methods.

METHODS

A total of 6 community detection methods are discussed: Walktrap, Edge-Betweenness, Infomap, Louvain, Label Propagation, and Spinglass. The Question Alignment approach is described and demonstrated using real-world data collected in 2013. This hypothetical case study was conducted in 2019 and focused on targeting a hand hygiene intervention to high-risk communities to prevent influenza transmission.

RESULTS

Community detection using the Walktrap method best fit the hypothetical case study. The communities derived using the Walktrap method were quite different from communities derived through the other 5 methods in both the number of communities and individuals within communities. There was evidence to support that the Question Alignment approach can help researchers produce more useful community detection results. Compared to other methods of selecting high-risk groups, the Walktrap produced the most communities that met the hypothetical intervention requirements.

CONCLUSIONS

As prevention research incorporating social networks increases, researchers can use the Question Alignment approach to produce more theoretically meaningful results and potentially more useful results for practice. Future research should focus on assessing whether the Question Alignment approach translates into improved intervention results.

Active Learning in an Environment of Innovative Training and Sustainability. Mapping of the Conceptual Structure of Research Fronts through a Bibliometric Analysis

The present study seeks to map and visualize up-to-date perspectives of the topic of active learning by analyzing and interpreting the different elements that make up learning ecosystems within the European Higher Education Area. With this aim, scientometric methods were employed to analyze a sample of 474 articles recovered from Web of Science (WoS) during the three-year period between 2018 and 2020. All articles examined the topic of active learning. Keywords (authors’ keywords and ‘keywords plus’) from the manuscripts were examined through co-occurrence analysis in order to establish the conceptual structure of active learning. Among the different trends and emerging topics identified, there is an important presence of topics related to technology applied to the field of education, where digital contexts acquire a preponderant role in current education. These innovative changes focused on the digital updating and exploitation of technology represent a methodological challenge that requires an involvement and commitment to this new space for educational practice by teachers and students.

Mapping flows on sparse networks with missing links

Unreliable network data can cause community-detection methods to overfit and highlight spurious structures with misleading information about the organization and function of complex systems. Here we show how to detect significant flow-based communities in sparse networks with missing links using the map equation. Since the map equation builds on Shannon entropy estimation, it assumes complete data such that analyzing undersampled networks can lead to overfitting. To overcome this problem, we incorporate a Bayesian approach with assumptions about network uncertainties into the map equation framework. Results in both synthetic and real-world networks show that the Bayesian estimate of the map equation provides a principled approach to revealing significant structures in undersampled networks.

The dynamics of negative stereotypes as revealed by tweeting behavior in the aftermath of the Charlie Hebdo terrorist attack

We describe the evolution of a stereotype as it emerged in tweets about the Charlie Hebdo terrorist attack in Paris in early 2015. Our focus is on terms associated with the Muslim community and the Islamic world. The data (400k tweets) were collected via Twitter streaming API and consisted of tweets that contained at least one of 16 hashtags associated with the Charlie Hebdo attack (e.g., #JeSuisCharlie, #IAmCharlie, #ParisAttacks), collected between January 14th and February 9th. From these data, we generated pairwise co-occurrence frequencies between key words such as “Islam”, “Muslim(s)”, “Arab(s)”, and “The Prophet” and possible associates such as: “terrorism”, “terror”, “terrorist(s)”, “kill(ed)”, “free”, “freedom” and “love”. We use changes in frequency of co-occurring words to define ways in which acute negative and positive stereotypes towards Muslims and Islam arise and evolve in three phases during the period of interest. We identify a positively-valenced backlash in a subset of tweets associated with the “origins of Islam”. Results depict the emergence and transformation of implicit online stereotypes related to Islam from naturally occurring social media data and how pro-as well as anti-Islam online small-world networks evolve in response to a terrorist attack.

Embedding-based Silhouette community detection

Mining complex data in the form of networks is of increasing interest in many scientific disciplines. Network communities correspond to densely connected subnetworks, and often represent key functional parts of real-world systems. This paper proposes the embedding-based Silhouette community detection (SCD), an approach for detecting communities, based on clustering of network node embeddings, i.e. real valued representations of nodes derived from their neighborhoods. We investigate the performance of the proposed SCD approach on 234 synthetic networks, as well as on a real-life social network. Even though SCD is not based on any form of modularity optimization, it performs comparably or better than state-of-the-art community detection algorithms, such as the InfoMap and Louvain. Further, we demonstrate that SCD's outputs can be used along with domain ontologies in semantic subgroup discovery, yielding human-understandable explanations of communities detected in a real-life protein interaction network. Being embedding-based, SCD is widely applicable and can be tested out-of-the-box as part of many existing network learning and exploration pipelines.

Functional network analysis reveals an immune tolerance mechanism in cancer

A major problem in data science is representation of data so that the variables driving key functions can be uncovered and explored. Correlation analysis is widely used to simplify networks of feature variables by reducing redundancies, but makes limited use of the network topology, relying on comparison of direct neighbor variables. The proposed method incorporates relational or functional profiles of neighboring variables along multiple common neighbors, which are fitted with Gaussian mixture models and compared using a data metric based on a version of optimal mass transport tailored to Gaussian mixtures. Hierarchical interactive visualization of the result leads to effective unbiased hypothesis generation. In a cancer gene expression study, this method uncovered an unanticipated immunosuppressive mechanism resembling maternal–fetal immune tolerance.

We present a technique to construct a simplification of a feature network which can be used for interactive data exploration, biological hypothesis generation, and the detection of communities or modules of cofunctional features. These are modules of features that are not necessarily correlated, but nevertheless exhibit common function in their network context as measured by similarity of relationships with neighboring features. In the case of genetic networks, traditional pathway analyses tend to assume that, ideally, all genes in a module exhibit very similar function, independent of relationships with other genes. The proposed technique explicitly relaxes this assumption by employing the comparison of relational profiles. For example, two genes which always activate a third gene are grouped together even if they never do so concurrently. They have common, but not identical, function. The comparison is driven by an average of a certain computationally efficient comparison metric between Gaussian mixture models. The method has its basis in the local connection structure of the network and the collection of joint distributions of the data associated with nodal neighborhoods. It is benchmarked on networks with known community structures. As the main application, we analyzed the gene regulatory network in lung adenocarcinoma, finding a cofunctional module of genes including the pregnancy-specific glycoproteins (PSGs). About 20% of patients with lung, breast, uterus, and colon cancer in The Cancer Genome Atlas (TCGA) have an elevated PSG+ signature, with associated poor group prognosis. In conjunction with previous results relating PSGs to tolerance in the immune system, these findings implicate the PSGs in a potential immune tolerance mechanism of cancers.

Allosteric Regulation at the Crossroads of New Technologies: Multiscale Modeling, Networks, and Machine Learning

Allosteric regulation is a common mechanism employed by complex biomolecular systems for regulation of activity and adaptability in the cellular environment, serving as an effective molecular tool for cellular communication. As an intrinsic but elusive property, allostery is a ubiquitous phenomenon where binding or disturbing of a distal site in a protein can functionally control its activity and is considered as the "second secret of life." The fundamental biological importance and complexity of these processes require a multi-faceted platform of synergistically integrated approaches for prediction and characterization of allosteric functional states, atomistic reconstruction of allosteric regulatory mechanisms and discovery of allosteric modulators. The unifying theme and overarching goal of allosteric regulation studies in recent years have been integration between emerging experiment and computational approaches and technologies to advance quantitative characterization of allosteric mechanisms in proteins. Despite significant advances, the quantitative characterization and reliable prediction of functional allosteric states, interactions, and mechanisms continue to present highly challenging problems in the field. In this review, we discuss simulation-based multiscale approaches, experiment-informed Markovian models, and network modeling of allostery and information-theoretical approaches that can describe the thermodynamics and hierarchy allosteric states and the molecular basis of allosteric mechanisms. The wealth of structural and functional information along with diversity and complexity of allosteric mechanisms in therapeutically important protein families have provided a well-suited platform for development of data-driven research strategies. Data-centric integration of chemistry, biology and computer science using artificial intelligence technologies has gained a significant momentum and at the forefront of many cross-disciplinary efforts. We discuss new developments in the machine learning field and the emergence of deep learning and deep reinforcement learning applications in modeling of molecular mechanisms and allosteric proteins. The experiment-guided integrated approaches empowered by recent advances in multiscale modeling, network science, and machine learning can lead to more reliable prediction of allosteric regulatory mechanisms and discovery of allosteric modulators for therapeutically important protein targets.

Exploring Evolution of Dynamic Networks via Diachronic Node Embeddings

Dynamic networks evolve with their structures changing over time. It is still a challenging problem to efficiently explore the evolution of dynamic networks in terms of both their structural and temporal properties. In this paper, we propose a visual analytics methodology to interactively explore the temporal evolution of dynamic networks in the context of their structure. A novel diachronic node embedding method is first proposed to learn latent representations of the structural and temporal features of nodes in a vector space. Diachronic node embeddings are then used to discover communities with similar structural proximity and temporal evolution patterns. A visual analytics system is designed to enable users to visually explore the evolutions of nodes, communities, and the network as a whole in terms of their structural and temporal properties. We evaluate the effectiveness of our method using artificial and real-world dynamic networks and comparisons with previous methods.

Brain network similarity: methods and applications

Graph theoretical approach has proved an effective tool to understand, characterize, and quantify the complex brain network. However, much less attention has been paid to methods that quantitatively compare two graphs, a crucial issue in the context of brain networks. Comparing brain networks is indeed mandatory in several network neuroscience applications. Here, we discuss the current state of the art, challenges, and a collection of analysis tools that have been developed in recent years to compare brain networks. We first introduce the graph similarity problem in brain network application. We then describe the methodological background of the available metrics and algorithms of comparing graphs, their strengths, and limitations. We also report results obtained in concrete applications from normal brain networks. More precisely, we show the potential use of brain network similarity to build a "network of networks" that may give new insights into the object categorization in the human brain. Additionally, we discuss future directions in terms of network similarity methods and applications.

Constructing and Cleaning Identity Graphs in the LOD Cloud

In the absence of a central naming authority on the Semantic Web, it is common for different data sets to refer to the same thing by different names. Whenever multiple names are used to denote the same thing, owl:sameAs statements are needed in order to link the data and foster reuse. Studies that date back as far as 2009, observed that the owl:sameAs property is sometimes used incorrectly. In our previous work, we presented an identity graph containing over 500 million explicit and 35 billion implied owl:sameAs statements, and presented a scalable approach for automatically calculating an error degree for each identity statement. In this paper, we generate subgraphs of the overall identity graph that correspond to certain error degrees. We show that even though the Semantic Web contains many erroneous owl:sameAs statements, it is still possible to use Semantic Web data while at the same time minimising the adverse effects of misusing owl:sameAs.

The Prospective Sense of Agency is Rooted in Local and Global Properties of Intrinsic Functional Brain Networks

The sense of agency (SoA) refers to a constitutional aspect of the self describing the extent to which individuals feel in control over their actions and consequences thereof. Although the SoA has been associated with mental health and well-being, it is still unknown how interindividual variability in the SoA is embedded in the intrinsic brain organization. We hypothesized that the prospective component of an implicit SoA is associated with brain networks related to SoA and sensorimotor predictions on multiple spatial scales. We replicated previous findings by showing a significant prospective SoA as indicated by intentional binding effects. Then, using task-free fMRI and graph analysis, we analyzed associations between intentional binding effects and the intrinsic brain organization at regional, modular, and whole-brain scales. The results showed that intermodular connections of a frontoparietal module including the premotor cortex, supramarginal gyrus, and dorsal precuneus are associated with individual differences in prospective intentional binding. Notably, prospective intentional binding effects were also related to global brain modularity within a specific structural resolution range. These findings suggest that an implicit SoA generated through sensorimotor predictions relies on the intrinsic organization of the brain connectome on both local and global scales.

A Relational Approach to Studying Collective Action in Dairy Cooperatives Producing Mountain Cheeses in the Alps: The Case of the Primiero Cooperative in the Eastern Italians Alps

Compared with more productive areas, mountain areas are at risk of being marginalized, particularly in the agri-food sector. To circumvent price competition, local actors in the mountains can develop specialized local products, which depends on their capacity to act collectively. Collective action, however, is complex and needs to be better understood if it is to steer initiatives towards success. This article sets out a relational approach to studying collective action in a dairy cooperative located in a mountain area: The Primiero cooperative in the Italian Alps. The common pool resources and territorial proximity frameworks were combined in a social network analysis of advice interactions among producer members, and an analysis of trust and conflict among members and between members and other actors involved in the value chain. The results show that the success of collective action can be explained by various complementary factors. Firstly, members had dense relationships, with high levels of trust and reciprocity, while the president had the role of prestige-based leader. Nonetheless, the analysis also highlighted conflicts related to the production levels of “traditional” and “intensive” producers, although members demonstrated a high capacity to resolve conflicts by creating their own rules to control further intensification. Socio-economic status did not appear to play a role in advice relationships, showing that the members interact horizontally. However, the results show that the geographical isolation of some members tended to inhibit their commitment to the collective dynamics. At a higher level, trust toward other actors involved in the value chain plays a central role in carrying out joint projects to develop and promote cheese.

Evaluating Nodes of Latent Mediators in Heterogeneous Communities

Conventionally, the importance of nodes in a network has been debated from the viewpoint of the amount of information received by the nodes and its neighbors. While node evaluation based on the adjacency relationship mainly uses local proximity information, the community structure that characterizes the network has hardly been considered. In this study, we propose a new node index that contributes to the understanding of the inter-community structure of a network by combining the measures of link distribution and community relevance. The visualization of node rankings and rank correlations with respect to the attack tolerance of networks demonstrated that the proposed index shows the highest performance in comparison with five previously proposed indexes, suggesting a new way to detect latent mediators in heterogeneous networks.

High dimensional model representation of log-likelihood ratio: binary classification with expression data

Background

Binary classification rules based on a small-sample of high-dimensional data (for instance, gene expression data) are ubiquitous in modern bioinformatics. Constructing such classifiers is challenging due to (a) the complex nature of underlying biological traits, such as gene interactions, and (b) the need for highly interpretable glass-box models. We use the theory of high dimensional model representation (HDMR) to build interpretable low dimensional approximations of the log-likelihood ratio accounting for the effects of each individual gene as well as gene-gene interactions. We propose two algorithms approximating the second order HDMR expansion, and a hypothesis test based on the HDMR formulation to identify significantly dysregulated pairwise interactions. The theory is seen as flexible and requiring only a mild set of assumptions.

Results

We apply our approach to gene expression data from both synthetic and real (breast and lung cancer) datasets comparing it also against several popular state-of-the-art methods. The analyses suggest the proposed algorithms can be used to obtain interpretable prediction rules with high prediction accuracies and to successfully extract significantly dysregulated gene-gene interactions from the data. They also compare favorably against their competitors across multiple synthetic data scenarios.

Conclusion

The proposed HDMR-based approach appears to produce a reliable classifier that additionally allows one to describe how individual genes or gene-gene interactions affect classification decisions. Both real and synthetic data analyses suggest that our methods can be used to identify gene networks with dysregulated pairwise interactions, and are therefore appropriate for differential networks analysis.

Social Sensing of the Imbalance of Urban and Regional Development in China Through the Population Migration Network around Spring Festival

Regional development differences are a universal problem in the economic development process of countries around the world. In recent decades, China has experienced rapid urban development since the implementation of the reform and opening-up policy. However, development differs across regions, triggering the migration of laborers from underdeveloped areas to developed areas. The interaction between regional development differences and Spring Festival has formed the world’s largest cyclical migration phenomenon, Spring Festival travel. Studying the migration pattern from public spatiotemporal behavior can contribute to understanding the differences in regional development. This paper proposes a geospatial network analytical framework to quantitatively characterize the imbalance of urban/regional development based on Spring Festival travel from the perspectives of complex network science and geospatial science. Firstly, the urban development difference is explored based on the intercity population flow difference ratio, PageRank algorithm, and attractiveness index. Secondly, the community detection method and rich-club coefficient are applied to further observe the spatial interactions between cities. Finally, the regional importance index and attractiveness index are used to reveal the regional development imbalance. The methods and findings can be used for urban planning, poverty alleviation, and population studies.

Urban Network and Regions in China: An Analysis of Daily Migration with Complex Networks Model

This paper analyzed urban network and regions in China using a complex network model. Data of daily migration among 348 prefectural-level cities from the Baidu Map location-based service (LBS) Open Platform were used to calculate urban network metrics and to delineate boundaries of urban regions. Results show that urban network in China displays an obvious hierarchy in terms of attracting and distributing population and controlling regional interaction. Regional integration has become increasingly prominent, as administrative boundaries and natural barriers no longer have strong impacts on urban connections. Overall, 18 urban regions were identified according to urban connectivity, and the degree of urban connection is higher among cities in the same urban region. Due to geographical proximity and close interaction, several provincial capital cities form an urban region with cities from neighboring provinces instead of those from the same province. Identification of urban region boundaries is of significant importance for sustainable development and policymaking on the demarcation of urban economic zones, urban agglomerations, and future adjustment of provincial administrative boundaries in China.

A Study on Dynamic Patterns of Technology Convergence with IPC Co-Occurrence-Based Analysis: The Case of 3D Printing

Technology convergence has become a typical characteristic of innovation, which affects the evolution of industrial structures and the core competitiveness of organizations. However, the existing research has mainly focused on the development of core areas of convergence, ignoring the potential breakthroughs that emerging peripheral convergence may bring. Therefore, this research put forward a comprehensive methodology based on IPC (International Patent Classification) co-occurrence analysis to study the dynamic patterns of technology convergence from the perspectives of reinforcing convergence and novel convergence. For the former, convergence trends in each period were explored by using association rules, and the convergence degree was measured based on the number of patents containing different IPC codes. Then, the corresponding core technical fields were identified by using information entropy. For the latter, a community detection algorithm based on IPC co-occurrence network was adopted to investigate the convergence trend by period, and important technology fields were identified by the centrality indicators. The methodology proposed in this study is beneficial for firms to seize technological opportunities in technology convergence.

Dimension Reduction and Clustering Models for Single-Cell RNA Sequencing Data: A Comparative Study

With recent advances in single-cell RNA sequencing, enormous transcriptome datasets have been generated. These datasets have furthered our understanding of cellular heterogeneity and its underlying mechanisms in homogeneous populations. Single-cell RNA sequencing (scRNA-seq) data clustering can group cells belonging to the same cell type based on patterns embedded in gene expression. However, scRNA-seq data are high-dimensional, noisy, and sparse, owing to the limitation of existing scRNA-seq technologies. Traditional clustering methods are not effective and efficient for high-dimensional and sparse matrix computations. Therefore, several dimension reduction methods have been introduced. To validate a reliable and standard research routine, we conducted a comprehensive review and evaluation of four classical dimension reduction methods and five clustering models. Four experiments were progressively performed on two large scRNA-seq datasets using 20 models. Results showed that the feature selection method contributed positively to high-dimensional and sparse scRNA-seq data. Moreover, feature-extraction methods were able to promote clustering performance, although this was not eternally immutable. Independent component analysis (ICA) performed well in those small compressed feature spaces, whereas principal component analysis was steadier than all the other feature-extraction methods. In addition, ICA was not ideal for fuzzy C-means clustering in scRNA-seq data analysis. K-means clustering was combined with feature-extraction methods to achieve good results.

Structural Characteristics and Spatial Patterns of the Technology Transfer Network in the Guangdong–Hong Kong–Macao Greater Bay Area

Recently, the Chinese government released the Outline of the Development Plan for the Guangdong–Hong Kong–Macao Greater Bay Area (GBA), raising the development of the GBA urban agglomeration to a national strategy. An efficient technology transfer network is conducive to promoting the integrated and coordinated development and enhancing the scientific and technological innovation capabilities of the GBA urban agglomeration. Therefore, this study uses the patent transaction data for three years (2010, 2014, and 2018), integrates data mining, and uses complex network analysis, based on full-flow and net-flow networks, from the overall characteristics, network node strength, network association, network node importance, and network communities to reveal the structural characteristics and spatial patterns of the technology transfer network in the GBA. The results revealed that: (1) Technology transfer networks (full-flow and net-flow) in the GBA show heterogeneity. (2) Full-flow network presents a clear hierarchy within the GBA, showing a “two poles and two strong” pattern, and technology transfer has the same city preference; outside the GBA, there are close technology transfer regions that have technical and geographical proximity characteristics; the net-flow network presents a “one pole, two strong” pattern, and Guangzhou has become the core region of the net-flow network. (3) Connection objects of the technology transfer network have path dependence and spatial preference. Coexistence of concentration and decentralization characterizes the spatial flow. (4) Spatial distribution of the hub and authority of the technology transfer network is heterogeneous and hierarchical. Each city in the GBA has its own technological advantages. (5) Spatial clustering characteristics of the community within the technology transfer network are obvious. (6) The GBA is dominated by the transfer of patented technology in the high-tech industry, while the transfer of patented technology in the traditional manufacturing industry also plays an important role.

Density-based Approach with Dual Optimization for Tracking Community Structure of Increasing Social Networks

The rapid evolution of social networks in recent years has focused the attention of researchers to find adequate solutions for the management of these networks. For this purpose, several efficient algorithms dedicated to the tracking and the rapid detection of the community structure have been proposed. In this paper, we propose a novel density-based approach with dual optimization for tracking community structure of increasing social networks. These networks are part of dynamic networks evolving by adding nodes with their links. The local optimization of the density makes it possible to reduce the resolution limit problem generated by the optimization of the modularity. The presented algorithm is incremental with a relatively low algorithmic complexity, making it efficient and faster. To demonstrate the effectiveness of our method, we test it on social networks of the real world. The experimental results show the performance and efficiency of our algorithm measured in terms of modularity density, modularity, normalized mutual information, number of communities discovered, running time and stability of communities.

Brain structural covariance networks in obsessive-compulsive disorder: a graph analysis from the ENIGMA Consortium

Brain structural covariance networks reflect covariation in morphology of different brain areas and are thought to reflect common trajectories in brain development and maturation. Large-scale investigation of structural covariance networks in obsessive-compulsive disorder (OCD) may provide clues to the pathophysiology of this neurodevelopmental disorder. Using T1-weighted MRI scans acquired from 1616 individuals with OCD and 1463 healthy controls across 37 datasets participating in the ENIGMA-OCD Working Group, we calculated intra-individual brain structural covariance networks (using the bilaterally-averaged values of 33 cortical surface areas, 33 cortical thickness values, and six subcortical volumes), in which edge weights were proportional to the similarity between two brain morphological features in terms of deviation from healthy controls (i.e. z-score transformed). Global networks were characterized using measures of network segregation (clustering and modularity), network integration (global efficiency), and their balance (small-worldness), and their community membership was assessed. Hub profiling of regional networks was undertaken using measures of betweenness, closeness, and eigenvector centrality. Individually calculated network measures were integrated across the 37 datasets using a meta-analytical approach. These network measures were summated across the network density range of K = 0.10-0.25 per participant, and were integrated across the 37 datasets using a meta-analytical approach. Compared with healthy controls, at a global level, the structural covariance networks of OCD showed lower clustering (P < 0.0001), lower modularity (P < 0.0001), and lower small-worldness (P = 0.017). Detection of community membership emphasized lower network segregation in OCD compared to healthy controls. At the regional level, there were lower (rank-transformed) centrality values in OCD for volume of caudate nucleus and thalamus, and surface area of paracentral cortex, indicative of altered distribution of brain hubs. Centrality of cingulate and orbito-frontal as well as other brain areas was associated with OCD illness duration, suggesting greater involvement of these brain areas with illness chronicity. In summary, the findings of this study, the largest brain structural covariance study of OCD to date, point to a less segregated organization of structural covariance networks in OCD, and reorganization of brain hubs. The segregation findings suggest a possible signature of altered brain morphometry in OCD, while the hub findings point to OCD-related alterations in trajectories of brain development and maturation, particularly in cingulate and orbitofrontal regions.

Abstract art paintings, global image properties, and verbal descriptions: An empirical and computational investigation

While global image properties (GIPs) relate to preference ratings in many categories of visual stimuli, this relationship is typically not seen for abstract art paintings. Using computational network science and empirical methods, we further investigated GIPs and subjective preferences. First, we replicated the earlier observation that GIPs do not relate to preferences for abstract art. Next, we estimated the network structure of abstract art paintings using two approaches: the first was based on verbal descriptions and the second on GIPs. We examined the extent to which network measures computed from these two networks (1) related to preference for abstract art paintings and (2) determined affiliation of images to specific art styles. Only semantic-based network predicted the subjective preference ratings and art style. Finally, preference and GIPs differed for sub-groups of abstract art paintings. Our results demonstrate the importance of verbal descriptors in evaluating abstract art, and that it is not useful in empirical aesthetics to treat abstract art paintings as a single category.

Assessment of congruence between co-occurrence and functional networks: A new framework for revealing community assembly rules

Describing how communities change over space and time is crucial to better understand and predict the functioning of ecosystems. We propose a new methodological framework, based on network theory and modularity concept, to determine which type of mechanisms (i.e. deterministic versus stochastic processes) has the strongest influence on structuring communities. This framework is based on the computation and comparison of two networks: the co-occurrence (based on species abundances) and the functional networks (based on the species traits values). In this way we can assess whether the species belonging to a given functional group also belong to the same co-occurrence group. We adapted the Dg index of Gauzens et al. (2015) to analyze congruence between both networks. This offers the opportunity to identify which assembly rule(s) play(s) the major role in structuring the community. We illustrate our framework with two datasets corresponding to different faunal groups and ecosystems, and characterized by different scales (spatial and temporal scales). By considering both species abundance and multiple functional traits, our framework improves significantly the ability to discriminate the main assembly rules structuring the communities. This point is critical not only to understand community structuring but also its response to global changes and other disturbances.

Determining the Hierarchical Architecture of the Human Brain Using Subject-Level Clustering of Functional Networks

Optimal integration and segregation of neuronal connections are necessary for efficient large-scale network communication between distributed cortical regions while allowing for modular specialization. This dynamic in the cortex is enabled at the network mesoscale by the organization of nodes into communities. Previous in vivo efforts to map the mesoscale architecture in humans had several limitations. Here we characterize a consensus multiscale community organization of the functional cortical network. We derive this consensus from the clustering of subject-level networks. We applied this analysis to magnetic resonance imaging data from 1003 healthy individuals part of the Human Connectome Project. The hierarchical atlas and code will be made publicly available for future investigators.

Latent cellular analysis robustly reveals subtle diversity in large-scale single-cell RNA-seq data

Single-cell RNA sequencing (scRNA-seq) is a powerful tool for characterizing the cell-to-cell variation and cellular dynamics in populations which appear homogeneous otherwise in basic and translational biological research. However, significant challenges arise in the analysis of scRNA-seq data, including the low signal-to-noise ratio with high data sparsity, potential batch effects, scalability problems when hundreds of thousands of cells are to be analyzed among others. The inherent complexities of scRNA-seq data and dynamic nature of cellular processes lead to suboptimal performance of many currently available algorithms, even for basic tasks such as identifying biologically meaningful heterogeneous subpopulations. In this study, we developed the Latent Cellular Analysis (LCA), a machine learning-based analytical pipeline that combines cosine-similarity measurement by latent cellular states with a graph-based clustering algorithm. LCA provides heuristic solutions for population number inference, dimension reduction, feature selection, and control of technical variations without explicit gene filtering. We show that LCA is robust, accurate, and powerful by comparison with multiple state-of-the-art computational methods when applied to large-scale real and simulated scRNA-seq data. Importantly, the ability of LCA to learn from representative subsets of the data provides scalability, thereby addressing a significant challenge posed by growing sample sizes in scRNA-seq data analysis.

Enhancing Personalized Recommendation by Implicit Preference Communities Modeling

Recommender systems aim to capture user preferences and provide accurate recommendations to users accordingly. For each user, there usually exist others with similar preferences, and a collection of users may also have similar preferences with each other, thus forming a community. However, such communities may not necessarily be explicitly given, and the users inside the same communities may not know each other; they are formally defined and named Implicit Preference Communities (IPCs) in this article. By enriching user preferences with the information of other users in the communities, the performance of recommender systems can also be enhanced.

Historical explicit ratings are a good resource to construct the IPCs of users but is usually sparse. Meanwhile, user preferences are easily affected by their social connections, which can be jointly used for IPC modeling with the ratings. However, this imposes two challenges for model design. First, the rating and social domains are heterogeneous; thus, it is challenging to coordinate social information and rating behaviors for a same learning task. Therefore, transfer learning is a good strategy for IPC modeling. Second, the communities are not explicitly labeled, and existing supervised learning approaches do not fit the requirement of IPC modeling. As co-clustering is an effective unsupervised learning approach for discovering block structures in high-dimensional data, it is a cornerstone for discovering the structure of IPCs.

In this article, we propose a recommendation model with Implicit Preference Communities from user ratings and social connections. To tackle the unsupervised learning limitation, we design a Bayesian probabilistic graphical model to capture the IPC structure for recommendation. Meanwhile, following the spirit of transfer learning, both rating behaviors and social connections are introduced into the model by parameter sharing. Moreover, Gibbs sampling-based algorithms are proposed for parameter inferences of the models. Furthermore, to meet the need for online scenarios when the data arrive sequentially as a stream, a novel online sampling-based parameter inference algorithm for recommendation is proposed. To the best of our knowledge, this is the first attempt to propose and formally define the concept of IPC.

Identifying genuine protein-protein interactions within communities of gene co-expression networks using a deconvolution method

Direct relationships between biological molecules connected in a gene co-expression network tend to reflect real biological activities such as gene regulation, protein-protein interactions (PPIs), and metabolisation. As correlation-based networks contain numerous indirect connections, those direct relationships are always 'hidden' in them. Compared with the global network, network communities imply more biological significance on predicting protein function, detecting protein complexes and studying network evolution. Therefore, identifying direct relationships in communities is a pervasive and important topic in the biological sciences. Unfortunately, this field has not been well studied. A major thrust of this study is to apply a deconvolution algorithm on communities stemming from different gene co-expression networks, which are constructed by fixing different thresholds for robustness analysis. Using the fifth Dialogue on Reverse Engineering Assessment and Methods challenge (DREAM5) framework, the authors demonstrate that nearly all new communities extracted from a 'deconvolution filter' contain more genuine PPIs than before deconvolution.

Exploring the Spatial Characteristics of Inbound Tourist Flows in China Using Geotagged Photos

As important modern tourist destinations, cities play a critical role in developing agglomerated tourism elements and promoting urban life quality. An in-depth exploration of tourist flow patterns between destination cities can reflect the dynamic trends of the inbound tourist market. This is significant for the development of tourism markets and innovation in tourism products. To this end, photos with geographical and corresponding metadata covering the entire country from 2011 to 2017 are used to explore the spatial characteristics of China’s inbound tourist flow, the spatial patterns of tourist movement, and the tourist destination cities group based on data mining techniques, including the Markov chain, a frequent-pattern-mining algorithm, and a community detection algorithm. Our findings show that: (1) the strongest flow of inbound tourists is between Beijing and Shanghai. These two cities, along with Xi’an and Guiling, form a “double-triangle” framework, (2) the travel between emerging destination cities in Central and Western China have gradually become frequently selected itineraries, and, (3) based on the flow intensity, inbound tourist destination cities can be divided into nine groups. This study provides a valuable reference for the development of China’s inbound tourism market.

Rewiring and indirect effects underpin modularity reshuffling in a marine food web under environmental shifts

Species are characterized by physiological and behavioral plasticity, which is part of their response to environmental shifts. Nonetheless, the collective response of ecological communities to environmental shifts cannot be predicted from the simple sum of individual species responses, since co-existing species are deeply entangled in interaction networks, such as food webs. For these reasons, the relation between environmental forcing and the structure of food webs is an open problem in ecology. To this respect, one of the main problems in community ecology is defining the role each species plays in shaping community structure, such as by promoting the subdivision of food webs in modules-that is, aggregates composed of species that more frequently interact-which are reported as community stabilizers. In this study, we investigated the relationship between species roles and network modularity under environmental shifts in a highly resolved food web, that is, a "weighted" ecological network reproducing carbon flows among marine planktonic species. Measuring network properties and estimating weighted modularity, we show that species have distinct roles, which differentially affect modularity and mediate structural modifications, such as modules reconfiguration, induced by environmental shifts. Specifically, short-term environmental changes impact the abundance of planktonic primary producers; this affects their consumers' behavior and cascades into the overall rearrangement of trophic links. Food web re-adjustments are both direct, through the rewiring of trophic-interaction networks, and indirect, with the reconfiguration of trophic cascades. Through such "systemic behavior," that is, the way the food web acts as a whole, defined by the interactions among its parts, the planktonic food web undergoes a substantial rewiring while keeping almost the same global flow to upper trophic levels, and energetic hierarchy is maintained despite environmental shifts. This behavior suggests the potentially high resilience of plankton networks, such as food webs, to dramatic environmental changes, such as those provoked by global change.