Modeling protective action decision-making in earthquakes by using explainable machine learning and video data.

Earthquakes pose substantial threats to communities worldwide. Understanding how people respond to the fast-changing environment during earthquakes is crucial for reducing risks and saving lives. This study aims to study people's protective action decision-making in earthquakes by leveraging explainable machine learning and video data. Specifically, this study first collected real-world CCTV footage and video postings from social media platforms, and then identified and annotated changes in the environment and people's behavioral responses during the M7.1 2018 Anchorage earthquake. By using the fully annotated video data, we applied XGBoost, a widely-used machine learning method, to model and forecast people's protective actions (e.g., drop and cover, hold on, and evacuate) during the earthquake. Then, explainable machine learning techniques were used to reveal the complex, nonlinear relationships between different factors and people's choices of protective actions. Modeling results confirm that social and environmental cues played critical roles in affecting the probability of different protective actions. Certain factors, such as the earthquake shaking intensity and number of people shown in the environment, displayed evident nonlinear relationships with the probability of choosing to evacuate. These findings can help emergency managers and policymakers design more effective protective action recommendations during earthquakes.

A highway vehicle routing dataset during the 2019 Kincade Fire evacuation.

As the threat of wildfire increases, it is imperative to enhance the understanding of household evacuation behavior and movements. Mobile GPS data provide a unique opportunity for studying evacuation routing behavior with high ecological validity, but there are little publicly available data. We generated a highway vehicle routing dataset derived from GPS trajectories generated by mobile devices (e.g., smartphones) in Sonoma County, California during the 2019 Kincade Fire that started on October 23, 2019. This dataset contains 21,160 highway vehicle routing records within Sonoma County from October 16, 2019 to November 13, 2019. The quality of the dataset is validated by checking trajectories and average travel speeds. The potential use of this dataset lies in analyzing and modeling evacuee route choice behavior, estimating traffic conditions during the evacuation, and validating wildfire evacuation simulation models.

Candidate gene prioritization for non-communicable diseases based on functional information: Case studies.

Candidate gene prioritization for complex non-communicable diseases is essential to understanding the mechanism and developing better means for diagnosing and treating these diseases. Many methods have been developed to prioritize candidate genes in protein-protein interaction (PPI) networks. Integrating functional information/similarity into disease-related PPI networks could improve the performance of prioritization. In this study, a candidate gene prioritization method was proposed for non-communicable diseases considering disease risks transferred between genes in weighted disease PPI networks with weights for nodes and edges based on functional information. Here, three types of non-communicable diseases with pathobiological similarity, Type 2 diabetes (T2D), coronary artery disease (CAD) and dilated cardiomyopathy (DCM), were used as case studies. Literature review and pathway enrichment analysis of top-ranked genes demonstrated the effectiveness of our method. Better performance was achieved after comparing our method with other existing methods. Pathobiological similarity among these three diseases was further investigated for common top-ranked genes to reveal their pathogenesis.

Prioritizing complex disease risk genes by integrating multiple data.

Complex diseases, such as obesity, type II diabetes and chronic obstructive pulmonary disease (COPD) as metabolic disorder-related diseases are major concern for worldwide public health in the 21st century. The identification of these disease risk genes has attracted increasing interest in computational systems biology. In this paper, a novel method was proposed to prioritize disease risk genes (PDRG) by integrating functional annotations, protein interactions and gene expression information to assess similarity between genes in a disease-related metabolic network. The gene prioritization method was successfully carried out for obesity and COPD, the effectiveness of which was superior to those of ToppGene and ToppNet in both literature validation and recall rate by LOOCV. Our method could be applied broadly to other metabolism-related diseases, helping to prioritize novel disease risk genes, and could shed light on diagnosis and effective therapies.

Assessing food system vulnerabilities: a fault tree modeling approach.

BACKGROUND: Food system function is vulnerable to disruption from a variety of sources. Disruption of the processes required for food provision may result in decreases in food security in affected communities. Currently, there are few tools that quantitatively predict or analyze food system vulnerabilities to contribute to food system resilience analysis. This work presents a prototype version of one such tool, a fault tree, which can be used conceptually and for future modeling work. Fault tree analysis is an engineering tool used to illustrate basic and intermediate factors that can cause overall system failures. METHODS: The fault tree defines food system functioning as food security at the community level and maps the components of the food system onto three main tenets of food security - accessibility, availability, and acceptability. Subtrees were populated using a top down approach guided by expertise, extant literature, and 36 stakeholder interviews. RESULTS: The food system is complex, requiring 12 subtrees to elaborate potential failures. Subtrees comprising accessibility include physical accessibility of the vending point and economic accessibility among community members. Food availability depends on the functioning of the food supply chain, or, in the case of individuals who rely on donated food, the food donation system. The food supply chain includes processing, wholesale operations, distribution systems, and retail center subtrees. Elements of acceptability include the medical appropriateness, nutritional adequacy, and cultural acceptability of food. Case studies of the effects of Winter Storm Jonas of 2016 and the 2013-2017 California drought in Baltimore City illustrate the utility of the fault tree model. CONCLUSION: FTA of potential routes to food system failure provides a tool that allows for consideration of the entirety of the food system; has potential to provide a quantitative assessment of food system failure and recovery; and is able to capture short-term and long-term hazards in a single framework. This systems modeling approach highlights an extensive list of vulnerability points throughout the food system, and underscores the message that reducing food system vulnerabilities requires action at all levels to protect communities from the risks of short-term and long-term threats to food security.

COPEWELL: A Conceptual Framework and System Dynamics Model for Predicting Community Functioning and Resilience After Disasters.

OBJECTIVE: Policy-makers and practitioners have a need to assess community resilience in disasters. Prior efforts conflated resilience with community functioning, combined resistance and recovery (the components of resilience), and relied on a static model for what is inherently a dynamic process. We sought to develop linked conceptual and computational models of community functioning and resilience after a disaster. METHODS: We developed a system dynamics computational model that predicts community functioning after a disaster. The computational model outputted the time course of community functioning before, during, and after a disaster, which was used to calculate resistance, recovery, and resilience for all US counties. RESULTS: The conceptual model explicitly separated resilience from community functioning and identified all key components for each, which were translated into a system dynamics computational model with connections and feedbacks. The components were represented by publicly available measures at the county level. Baseline community functioning, resistance, recovery, and resilience evidenced a range of values and geographic clustering, consistent with hypotheses based on the disaster literature. CONCLUSIONS: The work is transparent, motivates ongoing refinements, and identifies areas for improved measurements. After validation, such a model can be used to identify effective investments to enhance community resilience. (Disaster Med Public Health Preparedness. 2018;12:127-137).

Prioritizing chronic obstructive pulmonary disease (COPD) candidate genes in COPD-related networks.

Chronic obstructive pulmonary disease (COPD) is a multi-factor disease, which could be caused by many factors, including disturbances of metabolism and protein-protein interactions (PPIs). In this paper, a weighted COPD-related metabolic network and a weighted COPD-related PPI network were constructed base on COPD disease genes and functional information. Candidate genes in these weighted COPD-related networks were prioritized by making use of a gene prioritization method, respectively. Literature review and functional enrichment analysis of the top 100 genes in these two networks suggested the correlation of COPD and these genes. The performance of our gene prioritization method was superior to that of ToppGene and ToppNet for genes from the COPD-related metabolic network or the COPD-related PPI network after assessing using leave-one-out cross-validation, literature validation and functional enrichment analysis. The top-ranked genes prioritized from COPD-related metabolic and PPI networks could promote the better understanding about the molecular mechanism of this disease from different perspectives. The top 100 genes in COPD-related metabolic network or COPD-related PPI network might be potential markers for the diagnosis and treatment of COPD.

Chronic obstructive pulmonary disease candidate gene prioritization based on metabolic networks and functional information.

Chronic obstructive pulmonary disease (COPD) is a multi-factor disease, in which metabolic disturbances played important roles. In this paper, functional information was integrated into a COPD-related metabolic network to assess similarity between genes. Then a gene prioritization method was applied to the COPD-related metabolic network to prioritize COPD candidate genes. The gene prioritization method was superior to ToppGene and ToppNet in both literature validation and functional enrichment analysis. Top-ranked genes prioritized from the metabolic perspective with functional information could promote the better understanding about the molecular mechanism of this disease. Top 100 genes might be potential markers for diagnostic and effective therapies.