Individuals are motivated to increase their social status. To succeed in this pursuit, people must track information about others in their social sphere, monitor group norms, and adjust their behavior strategically. This study employed functional MRI and ecological momentary assessment methods in a sample of 92 college students belonging to 9 social groups to elucidate the neural mechanisms underlying these processes and their relationship to conformity in the context of alcohol use. When young adults passively looked at faces of their real-life social group peers, brain systems implicated in valuation and social cognition spontaneously tracked information about the popularity and leadership status of the social targets in an interdependent manner. Individual differences in these neural valuations were systematically associated with varying levels of conformity. Students who had stronger responses to faces of peers with relatively higher popularity and leadership status than themselves in one key valuation brain region, the ventromedial prefrontal cortex (vmPFC), were more likely to align their drinking behavior with their groups' norms in everyday life. These results provide evidence for how brain systems involved in valuation and social cognition flexibly track information about peers' popularity and leadership status in real-life social groups and contribute to a growing literature on the neural mechanisms through which social comparison processes shape conformity. Our study highlights the vmPFC as a central hub that spontaneously tracks status differences between the self and peers and uses this information to guide behavior to match group norms.

The commute time between two nodes in a network - the expected number of steps for a random walk to travel from one node to the other and then return - is a metric of broad importance arising in community detection, network routing, dimensionality reduction, and diffusion modeling. For random geometric graphs (RGGs), in which nodes are placed at random in a spatial domain and connected pairwise wherever their Euclidean distance is below a threshold radius, the relationship between commute times and the embedding geometry remains poorly understood outside very dense settings (where the role of the geometry disappears and commute times degenerate to a sum of inverse degrees). We develop and numerically validate a model for approximating commute times in sparse RGGs on a torus by combining theoretically motivated geometric contributions with an inverse degree sum. The geometric terms include a universal logarithmic contribution from the Laplacian, a quadratic correction encoding the compact topology of the torus, and a quartic angular term reflecting the square anisotropy of the domain. We fit this model to samples of node pairs across a range of graph sizes and mean degrees, demonstrating good predictive performance and that the geometric terms contribute significantly to model fit. We then study the continuous perturbation of the model from a regular square lattice to a fully random geometric graph, further validating the functional model form through this transition and showing how commute times in sparse RGGs retain meaningful geometric information about the embedding space.

Cognitive control is a suite of processes that helps individuals pursue goals despite resistance or uncertainty about what to do. Deficits of cognitive control underlie compulsive or risky behavior, as well as other clinical challenges associated with difficulties in regulating impulses, attention, thoughts, and feelings. Although cognitive control has been extensively studied as a dynamic feedback loop of perception, valuation, and action, it remains incompletely understood as a cohesive dynamic and distributed neural process. Here, we critically examine the history of and advances in the study of cognitive control, including how metaphors and cultural norms of power, morality, and rationality are intertwined with definitions of control, to consider holistically how different models explain which brain regions act as controllers. Controllers, the source of top-down signals, are typically localized in regions whose neural activations implement elementary component processes of control, including conflict monitoring and behavioral inhibition. Top-down signals from these regions guide the activation of other task-specific regions, biasing them towards task-specific activity patterns. A relatively new approach, network control theory, has roots in dynamical systems theory and systems engineering. This approach can mathematically show that controllers are regions with strongly nested and recurrent anatomical connectivity that efficiently propagate top-down signals, and precisely estimate the amount, location, and timing of signaling required to bias global activity to task-specific patterns. Importantly, the theory converges with established findings, provides new mathematical tools and intuitions for understanding control loops across levels of analysis, and naturally produces graded predictions of control across brain regions and modules of psychological function that have been unconsidered, marginalized, or indirectly linked. We describe how psychological and network control approaches converge and diverge, noting directions for future integration that could strengthen and sharpen our understanding and predictions of how the brain instantiates cognitive control.

The present study investigated between-person differences in daily positive emotion dynamics and their associations with flourishing across two studies (Study 1: n=244, Study 2: n=265). Three between-person indices of daily positive emotion dynamics were created: average intensity, variability, and inertia. Using latent profile analysis, a data-driven technique that identifies subgroups (referred to as profiles) within a population, four common ways in which these three emotion dynamics cluster at the person level were identified. Testing for associations between flourishing and the observed profiles of emotion dynamics revealed that people with high levels of positive emotion that were stable over time were highest in flourishing, followed by low-intensity but variable positive emotions, followed by individuals with low-intensity positive emotions. By considering how three key emotion dynamic indices cluster within individuals, we find that understanding both the average intensity and the extent of stability in daily positive emotion is necessary for understanding flourishing.

Modeling how networks change under structural perturbations can yield foundational insights into network robustness, which is critical in many real-world applications. The largest connected component is a popular measure of network performance. Percolation theory provides a theoretical framework to establish statistical properties of the largest connected component of large random graphs. However, this theoretical framework is typically only exact in the large- limit, failing to capture the statistical properties of largest connected components in small networks, which many real-world networks are. We derive expected values for the largest connected component of small G(,p) random graphs from which nodes are either removed uniformly at random or targeted by highest degree and compare these values with existing theory. We also visualize the performance of our expected values compared to existing theory for predicting the largest connected component of various real-world, small graphs.

Many empirical networks originate from correlational data, arising in domains as diverse as psychology, neuroscience, genomics, microbiology, finance, and climate science. Specialized algorithms and theory have been developed in different application domains for working with such networks, as well as in statistics, network science, and computer science, often with limited communication between practitioners in different fields. This leaves significant room for cross-pollination across disciplines. A central challenge is that it is not always clear how to best transform correlation matrix data into networks for the application at hand, and probably the most widespread method, i.e., thresholding on the correlation value to create either unweighted or weighted networks, suffers from multiple problems. In this article, we review various methods of constructing and analyzing correlation networks, ranging from thresholding and its improvements to weighted networks, regularization, dynamic correlation networks, threshold-free approaches, comparison with null models, and more. Finally, we propose and discuss recommended practices and a variety of key open questions currently confronting this field.

We propose a generative, mechanistic model of temporally evolving hypergraphs in which hyperedges form via noisy copying of previous hyperedges. Our proposed model reproduces several stylized facts from many empirical hypergraphs, is learnable from data, and defines a likelihood over a complete hypergraph rather than ego-based or other subhypergraphs. Analyzing our model, we derive asymptotic descriptions of the node degree, edge size, and edge intersection size distributions in terms of the model parameters. We also show several features of empirical hypergraphs which are and are not successfully captured by our model. We provide a scalable stochastic expectation maximization algorithm with which we can fit our model to hypergraph data sets with millions of nodes and edges. Finally, we assess our model on a hypergraph link prediction task, finding that an instantiation of our model with just 11 parameters can achieve competitive predictive performance with large neural networks.

Studies on college drinking-a behavior associated with health risks and reduced productivity-often rely on broadly defined peer or friendship networks. Yet, friendship networks can be further divided into more specific relational types, and the associations between such diverse social ties and drinking ties remain poorly understood. This study adopts a multilayer network framework to conduct a fine-grained examination of ten distinct types of social networks (i.e., layers)-including friendship, leadership, emotional support, and perceived drinking-by analyzing their structural similarity across three levels (node, link, and triad) and along a relationship-axis spectrum. We cluster networks based on these multi-level similarities, identifying three primary clusters (Affiliation, Leadership, and Drinking) and assess whether these structurally coherent clusters contribute to improved link prediction performance in perceived drinking networks. Both k-means clustering based on multi-level structural similarity and projection-based analysis along the hierarchical-horizontal spectrum revealed that perceived drinking nominations are structurally closest to leadership layers, followed by emotional support layers within the Affiliation cluster. Assessing whether these functional clusters can improve link prediction in perceived drinking networks, we find that layers within the same cluster yield predictive performance close to that of models using all layers. Notably, emotional support layers, which may reflect their structural proximity and distributed connectivity, offer the highest link prediction accuracy for drinking ties. Taken together, these findings demonstrate that coherent clusters of fine-grained, functionally and structurally aligned social layers facilitate more efficient inference in sparse or partially observed drinking-related networks. This, in turn, highlights their potential utility in predicting and preventing health-risk behaviors-such as alcohol use-that are typically difficult to observe or measure directly, and clarifies the types of relationships most strongly associated with such behaviors.

Influence propagation in social networks is a central problem in modern social network analysis, with important societal applications in politics and advertising. A large body of work has focused on cascading models, viral marketing, and finite-horizon diffusion. There is, however, a need for more developed, mathematically principled adversarial models, in which multiple, opposed actors strategically select nodes whose influence will maximally sway the crowd to their point of view. In the present work, we develop and analyze such a model based on harmonic functions and linear diffusion. Our general problem is known to be NP-hard and that the objective function is monotone and submodular; consequently, we can greedily approximate the solution within a constant factor. Introducing and analyzing a convex relaxation, we show that the problem can be approximately solved using smooth optimization methods. We illustrate the effectiveness of our approach on a variety of example networks.

Conversations shape future behaviors, particularly among young adults. However, young adults vary widely in their susceptibility to peer influence. What neural processes relate to this susceptibility? We examined whether activity in brain regions associated with social rewards and making sense of others’ minds relates to a common health behavior—drinking, following conversations about alcohol. We studied ten social groups of college students (N = 104 students; 4760 total observations) across two university campuses. We collected whole-brain fMRI data while participants viewed photographs of peers with whom they tended to drink at varying frequencies. Next, using ecological momentary assessment, we tracked alcohol conversations and drinking twice daily for 28 days. On average, talking about alcohol was associated with a higher likelihood of next-day drinking. Controlling for baseline drinking, participants who responded more strongly to peers with whom they drank alcohol more frequently—in brain regions associated with social rewards and mentalizing—showed a stronger, positive association between alcohol conversations and next-day drinking. Conversely, stronger neural responses to peers with whom they drank less frequently decoupled associations between alcohol conversations and next-day drinking. We conceptually replicate prior findings linking conversations and drinking in an observational, longitudinal setting and provide new evidence that neural responses to peers moderate links between alcohol conversations and drinking behavior among young adults.

The aging brain undergoes major changes in its topology. The mechanisms by which the brain mitigates age-associated changes in topology to maintain robust control of brain networks are unknown. Here we used diffusion MRI data from cognitively intact participants (n=480, ages 40-90) to study age-associated changes in the controllability of structural brain networks, features that could mitigate these changes, and the overall effect on cognitive function. We found age-associated declines in controllability in control hubs and large-scale networks, particularly within the and frontoparietal control and default mode networks. Redundancy, quantified via the assessment of multi-step paths within networks, mitigated the effects of changes in topology on network controllability. Lastly, network controllability, redundancy, and grey matter volume each played important complementary roles in cognitive function. In sum, our results highlight the importance of redundancy for robust control of brain networks and in cognitive function in healthy-aging.

Recent advancements in computational learning techniques have enabled the estimation of brain age (BA) from neuroimaging data. The difference between chronological age (CA) and BA, known as the BA gap, can potentially serve as a biomarker of brain health. Studies, however, have documented low correlations between BA gap and cognition in healthy aging. This suggests that protective mechanisms in the brain may help counter the effect of accelerated brain aging. Here, we investigated whether redundancy in brain networks may protect cognitive function in individuals with accelerated brain aging. First, we employed deep learning to estimate individual brain ages from structural magnetic resonance imaging (MRI). Next, we associated CA, BA, and BA gap, with cognitive measures and network topology derived from diffusion MRI and tractography. We found that CA and BA were both similarly related to cognitive measures and network topology, while BA gap did not show strong relationships in either domain. Despite observing no strong relationships between brain-age gap (BA gap) and demographic variables, cognitive measures, or topological features in healthy aging, individuals with accelerated aging (BA gap+) exhibited lower average degree and redundancy within the dorsal attention network compared to those with delayed aging (BA gap-). Furthermore, redundancy in the dorsal attention network was positively associated with processing speed in BA gap+ individuals. These results indicate a potential neuroprotective role of redundancy in structural brain networks for mitigating the impact of accelerated brain atrophy on cognitive performance in healthy aging.

We provide a rearrangement based algorithm for detection of subgraphs of k vertices with long escape times for directed or undirected networks that is not combinatorially complex to compute. Complementing other notions of densest subgraphs and graph cuts, our method is based on the mean hitting time required for a random walker to leave a designated set and hit the complement. We provide a new relaxation of this notion of hitting time on a given subgraph and use that relaxation to construct a subgraph detection algorithm that can be computed easily and a generalization to K -partitioning schemes. Using a modification of the subgraph detector on each component, we propose a graph partitioner that identifies regions where random walks live for comparably large times. Importantly, our method implicitly respects the directed nature of the data for directed graphs while also being applicable to undirected graphs. We apply the partitioning method for community detection to a large class of models and real -world data sets.

OBJECTIVES:Understanding disease transmission is a fundamental challenge in ecology. We used transmission potential networks to investigate whether a gastrointestinal protozoan (Blastocystis spp.) is spread through social, environmental, and/or zoonotic pathways in rural northeast Madagascar. MATERIALS AND METHODS:We obtained survey data, household GPS coordinates, and fecal samples from 804 participants. Surveys inquired about social contacts, agricultural activity, and sociodemographic characteristics. Fecal samples were screened for Blastocystis using DNA metabarcoding. We also tested 133 domesticated animals for Blastocystis. We used network autocorrelation models and permutation tests (network k-test) to determine whether networks reflecting different transmission pathways predicted infection. RESULTS:We identified six distinct Blastocystis subtypes among study participants and their domesticated animals. Among the 804 human participants, 74% (n = 598) were positive for at least one Blastocystis subtype. Close proximity to infected households was the most informative predictor of infection with any subtype (model averaged OR [95% CI]: 1.56 [1.33-1.82]), and spending free time with infected participants was not an informative predictor of infection (model averaged OR [95% CI]: 0.95 [0.82-1.10]). No human participant was infected with the same subtype as the domesticated animals they owned. DISCUSSION:Our findings suggest that Blastocystis is most likely spread through environmental pathways within villages, rather than through social or animal contact. The most likely mechanisms involve fecal contamination of the environment by infected individuals or shared food and water sources. These findings shed new light on human-pathogen ecology and mechanisms for reducing disease transmission in rural, low-income settings.

ABSTRACT The genome of Pseudomonas fluorescens encodes >50 proteins predicted to play a role in bis-(3′-5′)-cyclic dimeric guanosine monophosphate (c-di-GMP)-mediated biofilm formation. We built a network representation of protein–protein interactions and extracted key information via multidimensional scaling (i.e., principal component analysis) of node centrality measures, which measure features of proteins in a network. Proteins of different domain types (diguanylate cyclase, dual domain, phosphodiesterase, PilZ) exhibit unique network behavior and can be accurately classified by their network centrality values (i.e., roles in the network). The predictive power of protein–protein interactions in biofilm formation indicates the possibility of localized pools of c-di-GMP. A regression model showed a statistically significant impact of protein–protein interactions on the extent of biofilm formation in various environments. These results highlight the importance of a localized c-di-GMP signaling, extend our understanding of signaling by this second messenger beyond the current “Bow-tie Model,” support a newly proposed “Hub Model,” and suggest future avenues of investigation.

The use of network analysis as a tool has increased exponentially as more clinical researchers see the benefits of network data for modeling of infectious disease transmission or translational activities in a variety of areas, including patient-caregiving teams, provider networks, patient-support networks, and adoption of health behaviors or treatments, to name a few. Yet, relational data such as network data carry a higher risk of deductive disclosure. Cases of reidentification have occurred and this is expected to become more common as computational ability increases. Recent data sharing policies aim to promote reproducibility, support replicability, and protect federal investment in the effort to collect these research data by making them available for secondary analyses. However, typical practices to protect individual-level clinical research data may not be sufficiently protective of participant privacy in the case of network data, nor in some cases do they permit secondary data analysis. When sharing data, researchers must balance security, accessibility, reproducibility, and adaptability (suitability for secondary analyses). Here, we provide background about applying network analysis to health and clinical research, describe the pros and cons of applying typical practices for sharing clinical data to network data, and provide recommendations for sharing network data.

Despite the numerous ways now available to quantify which parts or subsystems of a network are most important, there remains a lack of centrality measures that are related to the complexity of information flows and are derived directly from entropy measures. Here, we introduce a ranking of edges based on how each edge's removal would change a system's von Neumann entropy (VNE), which is a spectral-entropy measure that has been adapted from quantum information theory to quantify the complexity of information dynamics over networks. We show that a direct calculation of such rankings is computationally inefficient (or unfeasible) for large networks: e.g.\ the scaling is $\mathcal{O}(N^3)$ per edge for networks with $N$ nodes. To overcome this limitation, we employ spectral perturbation theory to estimate VNE perturbations and derive an approximate edge-ranking algorithm that is accurate and fast to compute, scaling as $\mathcal{O}(N)$ per edge. Focusing on a form of VNE that is associated with a transport operator $e^{-\beta{ L}}$, where ${ L}$ is a graph Laplacian matrix and $\beta>0$ is a diffusion timescale parameter, we apply this approach to diverse applications including a network encoding polarized voting patterns of the 117th U.S. Senate, a multimodal transportation system including roads and metro lines in London, and a multiplex brain network encoding correlated human brain activity. Our experiments highlight situations where the edges that are considered to be most important for information diffusion complexity can dramatically change as one considers short, intermediate and long timescales $\beta$ for diffusion.

Real-world network datasets are typically obtained in ways that fail to capture all edges. The patterns of missing data are often non-uniform as they reflect biases and other shortcomings of different data collection methods. Nevertheless, uniform missing data is a common assumption made when no additional information is available about the underlying missing-edge pattern, and link prediction methods are frequently tested against uniformly missing edges. To investigate the impact of different missing-edge patterns on link prediction accuracy, we employ 9 link prediction algorithms from 4 different families to analyze 20 different missing-edge patterns that we categorize into 5 groups. Our comparative simulation study, spanning 250 real-world network datasets from 6 different domains, provides a detailed picture of the significant variations in the performance of different link prediction algorithms in these different settings. With this study, we aim to provide a guide for future researchers to help them select a link prediction algorithm that is well suited to their sampled network data, considering the data collection process and application domain.