Unequal impact and spatial aggregation distort COVID-19 growth rates.

The COVID-19 pandemic has posed unprecedented challenges to public health world-wide. To make decisions about mitigation strategies and to understand the disease dynamics, policy makers and epidemiologists must know how the disease is spreading in their communities. Here we analyse confirmed infections and deaths over multiple geographic scales to show that COVID-19's impact is highly unequal: many regions have nearly zero infections, while others are hot spots. We attribute the effect to a Reed-Hughes-like mechanism in which the disease arrives to regions at different times and grows exponentially at different rates. Faster growing regions correspond to hot spots that dominate spatially aggregated statistics, thereby skewing growth rates at larger spatial scales. Finally, we use these analyses to show that, across multiple spatial scales, the growth rate of COVID-19 has slowed down with each surge. These results demonstrate a trade-off when estimating growth rates: while spatial aggregation lowers noise, it can increase bias. Public policy and epidemic modelling should be aware of, and aim to address, this distortion. This article is part of the theme issue 'Data science approaches to infectious disease surveillance'.

Political Partisanship and Antiscience Attitudes in Online Discussions About COVID-19: Twitter Content Analysis.

BACKGROUND: The novel coronavirus pandemic continues to ravage communities across the United States. Opinion surveys identified the importance of political ideology in shaping perceptions of the pandemic and compliance with preventive measures. OBJECTIVE: The aim of this study was to measure political partisanship and antiscience attitudes in the discussions about the pandemic on social media, as well as their geographic and temporal distributions. METHODS: We analyzed a large set of tweets from Twitter related to the pandemic, collected between January and May 2020, and developed methods to classify the ideological alignment of users along the moderacy (hardline vs moderate), political (liberal vs conservative), and science (antiscience vs proscience) dimensions. RESULTS: We found a significant correlation in polarized views along the science and political dimensions. Moreover, politically moderate users were more aligned with proscience views, while hardline users were more aligned with antiscience views. Contrary to expectations, we did not find that polarization grew over time; instead, we saw increasing activity by moderate proscience users. We also show that antiscience conservatives in the United States tended to tweet from the southern and northwestern states, while antiscience moderates tended to tweet from the western states. The proportion of antiscience conservatives was found to correlate with COVID-19 cases. CONCLUSIONS: Our findings shed light on the multidimensional nature of polarization and the feasibility of tracking polarized opinions about the pandemic across time and space through social media data.

Poor attention: The wealth and regional gaps in event attention and coverage on Wikipedia.

Wikipedia is an important source of general knowledge covering a wide range of topics. Moreover, for many people around the world, it also serves as an essential news source for major events such as elections or disasters. Although Wikipedia covers many such events, some events are underrepresented and lack attention, despite their newsworthiness predicted from news value theory. In this paper, we analyze 17 490 event articles in four Wikipedia language editions and examine how the economic status and geographic region of the event location affects the attention and coverage it receives. We find that major Wikipedia language editions have a skewed focus, with more attention given to events in the world's more economically developed countries and less attention to events in less affluent regions. However, other factors, such as the number of deaths in a disaster, are also associated with the attention an event receives. Overall, this work provides a nuanced understanding of attention and coverage on Wikipedia through event articles and adds new empirical analysis to news value theory.

Clique densification in networks.

Real-world networks are rarely static. Recently, there has been increasing interest in both network growth and network densification, in which the number of edges scales superlinearly with the number of nodes. Less studied but equally important, however, are scaling laws of higher-order cliques, which can drive clustering and network redundancy. In this paper, we study how cliques grow with network size, by analyzing several empirical networks from emails to Wikipedia interactions. Our results show superlinear scaling laws whose exponents increase with clique size, in contrast to predictions from a previous model. We then show that these results are in qualitative agreement with a model that we propose, the local preferential attachment model, where an incoming node links not only to a target node, but also to its higher-degree neighbors. Our results provide insights into how networks grow and where network redundancy occurs.

What's On the Menu? Towards Predicting Nutritional Quality of Food Environments.

Unhealthy diets are a leading cause of major chronic diseases including obesity, diabetes, cancer, and heart disease. Food environments-the physical spaces in which people access and consume food-have the potential to profoundly impact diet and related diseases. We take a step towards better understanding the nutritional quality of food environments by developing MINT: Menu Item to NutrienT model. This model utilizes under-studied data sources on recipes and generic food items, along with state-of-the-art word embedding and deep learning methods, to predict the nutrient density of never-before-seen food items using only their name as input. The model achieves an R2=0.77, a substantial improvement over comparable models. We illustrate the utility of MINT by applying it to the Los Angeles restaurant food environment, and discover close agreement between predicted and ground truth nutrient density of restaurant menu items. This model represents a significant step towards a policy toolkit needed to precisely identify and target food environments characterized by poor nutritional quality.

Selecting Clustering Algorithms for Identity-By-Descent Mapping.

Groups of distantly related individuals who share a short segment of their genome identical-by-descent (IBD) can provide insights about rare traits and diseases in massive biobanks using IBD mapping. Clustering algorithms play an important role in finding these groups accurately and at scale. We set out to analyze the fitness of commonly used, fast and scalable clustering algorithms for IBD mapping applications. We designed a realistic benchmark for local IBD graphs and utilized it to compare the statistical power of clustering algorithms via simulating 2.3 million clusters across 850 experiments. We found Infomap and Markov Clustering (MCL) community detection methods to have high statistical power in most of the scenarios. They yield a 30% increase in power compared to the current state-of-art approach, with a 3 orders of magnitude lower runtime. We also found that standard clustering metrics, such as modularity, cannot predict statistical power of algorithms in IBD mapping applications. We extend our findings to real datasets by analyzing the Population Architecture using Genomics and Epidemiology (PAGE) Study dataset with 51,000 samples and 2 million shared segments on Chromosome 1, resulting in the extraction of 39 million local IBD clusters. We demonstrate the power of our approach by recovering signals of rare genetic variation in the Whole-Exome Sequence data of 200,000 individuals in the UK Biobank. We provide an efficient implementation to enable clustering at scale for IBD mapping for various populations and scenarios.Supplementary Information: The code, along with supplementary methods and figures are available at https://github.com/roohy/localIBDClustering.

Leveraging change point detection to discover natural experiments in data.

Change point detection has many practical applications, from anomaly detection in data to scene changes in robotics; however, finding changes in high dimensional data is an ongoing challenge. We describe a self-training model-agnostic framework to detect changes in arbitrarily complex data. The method consists of two steps. First, it labels data as before or after a candidate change point and trains a classifier to predict these labels. The accuracy of this classifier varies for different candidate change points. By modeling the accuracy change we can infer the true change point and fraction of data affected by the change (a proxy for detection confidence). We demonstrate how our framework can achieve low bias over a wide range of conditions and detect changes in high dimensional, noisy data more accurately than alternative methods. We use the framework to identify changes in real-world data and measure their effects using regression discontinuity designs, thereby uncovering potential natural experiments, such as the effect of pandemic lockdowns on air pollution and the effect of policy changes on performance and persistence in a learning platform. Our method opens new avenues for data-driven discovery due to its flexibility, accuracy and robustness in identifying changes in data.

The transsortative structure of networks.

Network topologies can be highly non-trivial, due to the complex underlying behaviours that form them. While past research has shown that some processes on networks may be characterized by local statistics describing nodes and their neighbours, such as degree assortativity, these quantities fail to capture important sources of variation in network structure. We define a property called transsortativity that describes correlations among a node's neighbours. Transsortativity can be systematically varied, independently of the network's degree distribution and assortativity. Moreover, it can significantly impact the spread of contagions as well as the perceptions of neighbours, known as the majority illusion. Our work improves our ability to create and analyse more realistic models of complex networks.

Inferring models of opinion dynamics from aggregated jury data.

Jury deliberations provide a quintessential example of collective decision-making, but few studies have probed the available data to explore how juries reach verdicts. We examine how features of jury dynamics can be better understood from the joint distribution of final votes and deliberation time. To do this, we fit several different decision-making models to jury datasets from different places and times. In our best-fit model, jurors influence each other and have an increasing tendency to stick to their opinion of the defendant's guilt or innocence. We also show that this model can explain spikes in mean deliberation times when juries are hung, sub-linear scaling between mean deliberation times and trial duration, and unexpected final vote and deliberation time distributions. Our findings suggest that both stubbornness and herding play an important role in collective decision-making, providing a nuanced insight into how juries reach verdicts, and more generally, how group decisions emerge.

Cascading failures in scale-free interdependent networks.

Large cascades are a common occurrence in many natural and engineered complex systems. In this paper we explore the propagation of cascades across networks using realistic network topologies, such as heterogeneous degree distributions, as well as intra- and interlayer degree correlations. We find that three properties, scale-free degree distribution, internal network assortativity, and cross-network hub-to-hub connections, are all necessary components to significantly reduce the size of large cascades in the Bak-Tang-Wiesenfeld sandpile model. We demonstrate that correlations present in the structure of the multilayer network influence the dynamical cascading process and can prevent failures from spreading across connected layers. These findings highlight the importance of internal and cross-network topology in optimizing robustness of interconnected systems.

Partial Shocks on Cooperative Multiplex Networks with Varying Degrees of Noise.

In many cooperative networks, such as alliance and trade networks, abrupt and intense changes to the state of the system (which we call "shocks"), can substantially change the network. We examine how such shocks affect multiplex networks via an agent-based model, in which agents add, drop, or change ties to increase their utility. At a certain time-point, some agents are "shocked" by changing (increasing or decreasing) the cost associated with tie-formation or tie-maintenance. Our model makes several improvements to previous models, including (a) only a fraction of nodes are shocked to simulate small wars or scattered tariff increases or decreases and (b) agents can make both utility-maximizing decisions and randomly rewire ties to explore the utility landscape. Interestingly, we find that randomly rewiring ties increases the utility of agents, for reasons similar to simulated annealing in physics. Furthermore, we create a novel metric to determine how networks change after a shock and find that the size of a shock and noise significantly changes the network, but only when agents' incentives for tie-formation are sufficiently high. Together, these results suggest that adding more realism to cooperation network models can give nuanced understanding to network shocks.

Self-organization of dragon king failures.

The mechanisms underlying cascading failures are often modeled via the paradigm of self-organized criticality. Here we introduce a simple network model where nodes self-organize to be either weakly or strongly protected against failure in a manner that captures the trade-off between degradation and reinforcement of nodes inherent in many network systems. If strong nodes cannot fail, any failure is contained to a single, isolated cluster of weak nodes and the model produces power-law distributions of failure sizes. We classify the large, rare events that involve the failure of only a single cluster as "black swans." In contrast, if strong nodes fail once a sufficient fraction of their neighbors fail, then failure can cascade across multiple clusters of weak nodes. If over 99.9% of the nodes fail due to this cluster hopping mechanism, we classify this as a "dragon king," which are massive failures caused by mechanisms distinct from smaller failures. The dragon kings observed are self-organized, existing over a wide range of reinforcement rates and system sizes. We find that once an initial cluster of failing weak nodes is above a critical size, the dragon king mechanism kicks in, leading to piggybacking system-wide failures. We demonstrate that the size of the initial failed weak cluster predicts the likelihood of a dragon king event with high accuracy and we develop a simple control strategy that can dramatically reduce dragon kings and other large failures.

The myopia of crowds: Cognitive load and collective evaluation of answers on Stack Exchange.

Crowds can often make better decisions than individuals or small groups of experts by leveraging their ability to aggregate diverse information. Question answering sites, such as Stack Exchange, rely on the "wisdom of crowds" effect to identify the best answers to questions asked by users. We analyze data from 250 communities on the Stack Exchange network to pinpoint factors affecting which answers are chosen as the best answers. Our results suggest that, rather than evaluate all available answers to a question, users rely on simple cognitive heuristics to choose an answer to vote for or accept. These cognitive heuristics are linked to an answer's salience, such as the order in which it is listed and how much screen space it occupies. While askers appear to depend on heuristics to a greater extent than voters when choosing an answer to accept as the most helpful one, voters use acceptance itself as a heuristic, and they are more likely to choose the answer after it has been accepted than before that answer was accepted. These heuristics become more important in explaining and predicting behavior as the number of available answers to a question increases. Our findings suggest that crowd judgments may become less reliable as the number of answers grows.

Competing opinions and stubborness: Connecting models to data.

We introduce a general contagionlike model for competing opinions that includes dynamic resistance to alternative opinions. We show that this model can describe candidate vote distributions, spatial vote correlations, and a slow approach to opinion consensus with sensible parameter values. These empirical properties of large group dynamics, previously understood using distinct models, may be different aspects of human behavior that can be captured by a more unified model, such as the one introduced in this paper.

Testing Modeling Assumptions in the West Africa Ebola Outbreak.

The Ebola virus in West Africa has infected almost 30,000 and killed over 11,000 people. Recent models of Ebola Virus Disease (EVD) have often made assumptions about how the disease spreads, such as uniform transmissibility and homogeneous mixing within a population. In this paper, we test whether these assumptions are necessarily correct, and offer simple solutions that may improve disease model accuracy. First, we use data and models of West African migration to show that EVD does not homogeneously mix, but spreads in a predictable manner. Next, we estimate the initial growth rate of EVD within country administrative divisions and find that it significantly decreases with population density. Finally, we test whether EVD strains have uniform transmissibility through a novel statistical test, and find that certain strains appear more often than expected by chance.