A systems approach to investigate GPCR-mediated Ras signaling network in chemoattractant sensing

A GPCR-mediated signaling network enables a chemotactic cell to generate adaptative Ras signaling in response to a large range of concentrations of a chemoattractant. To explore potential regulatory mechanisms of GPCR-controlled Ras signaling in chemosensing, we applied a software package, Simmune, to construct detailed spatiotemporal models simulating responses of the cAR1-mediated Ras signaling network. We first determined dynamics of G-protein activation and Ras signaling in Dictyostelium cells in response to cAMP stimulations using live-cell imaging and then constructed computation models by incorporating potential mechanisms. Using simulations, we validated the dynamics of signaling events and predicted the dynamic profiles of those events in the cAR1-mediated Ras signaling networks with defective Ras inhibitory mechanisms, such as without RasGAP, with RasGAP overexpression, or RasGAP hyperactivation. We described a method of using Simmune to construct spatiotemporal models of a signaling network and run computational simulations without writing mathematical equations. This approach will help biologists to develop and analyze computational models that parallel live-cell experiments.

Trends of Street Fights/Quarrels in Iran 2013 - 2018: A Bayesian Spatiotemporal Perspective

Background: Conflict/quarrel, as one of the indicators of violence, is a social issue still seen in all societies. It occurs between two or more people or groups in a social relationship and can disrupt society order and possesses destructive consequences for disputants and society. Objectives: The present study aimed to evaluate points and trends of relative risk (RR) of quarrels in Iran for total population and both sexes separately by using spatiotemporal models. Methods: Official data published by Iranian Legal Medicine Organization (ILMO) from 2013 to 2018 was studied. Spatiotemporal methods were used for analyzing the data and producing relevant maps. These models overcome the problems related to usual estimates of RR and are capable of covering spatial and temporal effects and their interactions simultaneously. Results: The results showed that Ardabil (P2, RR = 1.32), Chaharmahal and Bakhtiari, and Kohgiluyeh and Boyer-Ahmad (RR = 1.1 - 1.3) provinces had the highest risk of street quarrel for total population. The results for males are the same as the results for the total population. There was the highest risk for females in Alborz (P5, RR = 1.38) province. The risk was the lowest for the southern provinces of Iran for the total population (0.3 - 0.7), females (0.3 - 0.55), and for males (0.3-0.6). There was no significant change in RR over time for males and total population. However, there is an apparent decreasing trend for females. Conclusions: In general, southern parts of Iran have lower risk of street fights/quarrels. Street fight is a multifactor phenomenon that could leave various consequences on society. It seems necessary to conduct further research to find out the reasons for its occurrence in different parts of the country.

Hierarchical Gaussian Processes and Mixtures of Experts to Model COVID-19 Patient Trajectories

Gaussian processes (GPs) are a versatile nonparametric model for nonlinear regression and have been widely used to study spatiotemporal phenomena. However, standard GPs offer limited interpretability and generalizability for datasets with naturally occurring hierarchies. With large-scale, rapidly-updating electronic health record (EHR) data, we want to study patient trajectories across diverse patient cohorts while preserving patient subgroup structure. In this work, we partition our cohort of over 2000 COVID-19 patients by sex and ethnicity. We develop and apply a hierarchical Gaussian process and a mixture of experts (MOE) hierarchical GP model to fit patient trajectories on clinical markers of disease progression. A case study for albumin, an effective predictor of COVID-19 patient outcomes, highlights the predictive performance of these models. These hierarchical spatiotemporal models of EHR data bring us a step closer toward our goal of building flexible approaches to capture patient data that can be used in real-time systems.

Spatiotemporal tracing of pandemic spread from infection data

COVID-19, a global pandemic caused by the Severe Acute Respiratory Syndrome Coronavirus 2 virus, has claimed millions of lives worldwide. Amid soaring contagion due to newer strains of the virus, it is imperative to design dynamic, spatiotemporal models to contain the spread of infection during future outbreaks of the same or variants of the virus. The reliance on existing prediction and contact tracing approaches on prior knowledge of inter- or intra-zone mobility renders them impracticable. We present a spatiotemporal approach that employs a network inference approach with sliding time windows solely on the date and number of daily infection numbers of zones within a geographical region to generate temporal networks capturing the influence of each zone on another. It helps analyze the spatial interaction among the hotspot or spreader zones and highly affected zones based on the flow of network contagion traffic. We apply the proposed approach to the daily infection counts of New York State as well as the states of USA to show that it effectively measures the phase shifts in the pandemic timeline. It identifies the spreaders and affected zones at different time points and helps infer the trajectory of the pandemic spread across the country. A small set of zones periodically exhibit a very high outflow of contagion traffic over time, suggesting that they act as the key spreaders of infection. Moreover, the strong influence between the majority of non-neighbor regions suggests that the overall spread of infection is a result of the unavoidable long-distance trips by a large number of people as opposed to the shorter trips at a county level, thereby informing future mitigation measures and public policies.

Explainable Learning Based on Hyperparameters Optimization of Lightweight Spatiotemporal Models for Cellular Traffic Prediction

The rapid development of the Internet of Things and multimedia applications has led to an exponential growth in mobile network traffic year by year. In order to meet the demand for large amounts of data transmission and solve the problem of insufficient spectrum resources, millimeter waves are adopted for 5G communication. For B5G/6G, effective use of spectrum resources is one of the key technologies for the development of mobile communications. Therefore, this study uses a lightweight neural network to predict cellular traffic based on regional data, considering the data types of temporal and spatial dependence at the same time. Furthermore, in order to optimize the prediction performance and reduce the number of parameters of the neural network, this study uses a meta-heuristic algorithm to adjust the hyperparameters and combines local and global explanations to interpret the improvement of traffic prediction. The local explanations show the adjustment results of a single hyperparameter, and global explanations show the correlation between different hyperparameters and their influence on the amount and accuracy of model parameters. The simulation results show that compared with adjustment strategies of the manual method and greedy algorithm, the proposed explainable learning method can effectively improve the accuracy of cellular traffic prediction, reduce the number of parameters and provide a reasonable explanation.

Estimating spatiotemporal patterns of deaths by COVID-19 outbreak on a global scale

ObjectiveOur main objective is to estimate the trend of deaths by COVID-19 on a global scale, considering the six continents.Study designThe study design was a retrospective observational study conducted using the secondary data provided by the Our World in Data project on a public domain.SettingThis study was conducted based on worldwide deaths by COVID-19 recorded for the Our World in Data project from 29 February 2020 to 17 February 2021.MethodsEstimating the trend in COVID-19 deaths is not a trivial task due to the problems associated with the COVID-19 data, such as the spatial and temporal heterogeneity, observed seasonality and the delay between the onset of symptoms and diagnosis, indicating a relevant measurement error problem and changing the series’ dependency structure. To bypass the aforementioned problems, we propose a method to estimate the components of trend, seasonality and cycle in COVID-19 data, controlling for the presence of measurement error and considering the spatial heterogeneity. We used the proposed model to estimate the trend component of deaths by COVID-19 on a global scale.ResultsThe model was able to capture the patterns in the occurrence of deaths related to COVID-19, overcoming the problems observed in COVID-19 data. We found compelling evidence that spatiotemporal models are more accurate than univariate models to estimate the patterns of the occurrence of deaths. Based on the measures of dispersion of the models’ prediction in relation to observed deaths, it is possible to note that the models with spatial component are significantly superior to the univariate model.ConclusionThe findings suggested that the spatial dynamics have an important role in the COVID-19 epidemic process since the results provided evidence that spatiotemporal models are more accurate to estimate the general patterns of the occurrence of deaths related to COVID-19.

Features of statistical modeling and forecasting of crime: theoretical aspect

The purpose of the work is to provide a theoretical analysis of modern methods of modeling and forecasting the state of crime, which can be used in the system of public administration of the law enforcement sphere. In the course of the research, the peculiarities of using various tools and models for predicting the state of crime are revealed. A significant part of the research of scientists is directed towards the use of spatial and spatiotemporal models, as well as methods of artificial intelligence. The high quality of monthly forecasts is noted. Various economic, social, geographical, temporal and other groups of factors that influence the state of crime are identified. It is established that the quality of the developed crime forecasts depends on the choice of the optimal method and period of forecasting, on the completeness of the information base, including social, economic, legal and other characteristics of the phenomena and processes of public life that affect the criminal situation. It is noted that the practical use of artificial intelligence and econometric analysis methods in predicting the state of crime is becoming particularly relevant at the present time.

Classification of Brain Tumours in MR Images using Deep Spatiospatial Models

A brain tumour is a mass or cluster of abnormal cells in the brain, which has the possibility of becoming life-threatening because of its ability to invade neighbouring tissues and also form metastases. An accurate diagnosis is essential for successful treatment planning and magnetic resonance imaging is the principal imaging modality for diagnostic of brain tumours and their extent. Deep Learning methods in computer vision applications have shown significant improvement in recent years, most of which can be credited to the fact that a sizeable amount of data is available to train models on, and the improvements in the model architectures yielding better approximations in a supervised setting. Classifying tumours using such deep learning methods has made significant progress with the availability of open datasets with reliable annotations. Typically those methods are either 3D models, which use 3D volumetric MRIs or even 2D models considering each slice separately. However, by treating one spatial dimension separately or by considering the slices as a sequence of images over time, spatiotemporal models can be employed as "spatiospatial" models for this task. These models have the capabilities of learning specific spatial and temporal relationship, while reducing computational costs. This paper uses two spatiotemporal models, ResNet (2+1)D and ResNet Mixed Convolution, to classify different types of brain tumours. It was observed that both these models performed superior to the pure 3D convolutional model, ResNet18. Furthermore, it was also observed that pre-training the models on a different, even unrelated dataset before training them for the task of tumour classification improves the performance. Finally, Pre-trained ResNet Mixed Convolution was observed to be the best model in these experiments, achieving a macro F1-score of 0.9345 and a test accuracy of 96.98%, while at the same time being the model with the least computational cost.