Multiobjective grammar-based genetic programming applied to the study of asthma and allergy epidemiology

Spatial distribution of hospital admissions for asthma in the central area of Asturias, Northern Spain

BACKGROUND

Asturias is one of the communities with the highest rates of hospital admission for asthma in Spain. The environmental pollution or people lifestyle are some of the factors that contribute to the appearance or aggravation of this illness. The aim of this study was to show the spatial distribution of asthma admissions risks in the central municipalities of Asturias and to analyze the observed spatial patterns.

METHODS

Urgent hospital admissions for asthma and status asthmaticus occurred between 2016 to 2018 on the public hospitals of the central area of Asturias were used. Population data were assigned in 5 age groups. Standardised admission ratio (SAR), smoothed relative risk (SRR) and posterior risk probability (PP) were calculated for each census tract (CT). A spatial trend analysis was run, a spatial autocorrelation index (Morans I) was calculated and a cluster and outlier analysis (Anselin Local Morans I) was finally performed in order to analyze spatial clusters.

RESULTS

The total number of hospital urgent asthma admissions during the study period was 2324, 1475 (63.46%) men and 849 (36.56%) women. The municipalities with the highest values of SRR and PP were located on the northwest area: Avilés, Gozón, Carreño, Corvera de Asturias, Castrillón and Illas. A high risk cluster was found for the municipalities of Avilés, Gozón y Corvera de Asturias.

CONCLUSIONS

The spatial analysis showed high risk of hospitalization for asthma on the municipalities of the northwest area of the study, which highlight the existence of spatial inequalities on the distribution of urgent hospital admissions.

Post-Analysis of Predictive Modeling with an Epidemiological Example

Post-analysis of predictive models fosters their application in practice, as domain experts want to understand the logic behind them. In epidemiology, methods explaining sophisticated models facilitate the usage of up-to-date tools, especially in the high-dimensional predictor space. Investigating how model performance varies for subjects with different conditions is one of the important parts of post-analysis. This paper presents a model-independent approach for post-analysis, aiming to reveal those subjects' conditions that lead to low or high model performance, compared to the average level on the whole sample. Conditions of interest are presented in the form of rules generated by a multi-objective evolutionary algorithm (MOGA). In this study, Lasso logistic regression (LLR) was trained to predict cardiovascular death by 2016 using the data from the 1984-1989 examination within the Kuopio Ischemic Heart Disease Risk Factor Study (KIHD), which contained 2682 subjects and 950 preselected predictors. After 50 independent runs of five-fold cross-validation, the model performance collected for each subject was used to generate rules describing "easy" and "difficult" cases. LLR with 61 selected predictors, on average, achieved 72.53% accuracy on the whole sample. However, during post-analysis, three categories of subjects were discovered: "Easy" cases with an LLR accuracy of 95.84%, "difficult" cases with an LLR accuracy of 48.11%, and the remaining cases with an LLR accuracy of 71.00%. Moreover, the rule analysis showed that medication was one of the main confusing factors that led to lower model performance. The proposed approach provides insightful information about subjects' conditions that complicate predictive modeling.

Classification algorithm for congenital Zika Syndrome: characterizations, diagnosis and validation

Zika virus was responsible for the microcephaly epidemic in Brazil which began in October 2015 and brought great challenges to the scientific community and health professionals in terms of diagnosis and classification. Due to the difficulties in correctly identifying Zika cases, it is necessary to develop an automatic procedure to classify the probability of a CZS case from the clinical data. This work presents a machine learning algorithm capable of achieving this from structured and unstructured available data. The proposed algorithm reached 83% accuracy with textual information in medical records and image reports and 76% accuracy in classifying data without textual information. Therefore, the proposed algorithm has the potential to classify CZS cases in order to clarify the real effects of this epidemic, as well as to contribute to health surveillance in monitoring possible future epidemics.

Integrating Computational Methods to Investigate the Macroecology of Microbiomes

Studies in microbiology have long been mostly restricted to small spatial scales. However, recent technological advances, such as new sequencing methodologies, have ushered an era of large-scale sequencing of environmental DNA data from multiple biomes worldwide. These global datasets can now be used to explore long standing questions of microbial ecology. New methodological approaches and concepts are being developed to study such large-scale patterns in microbial communities, resulting in new perspectives that represent a significant advances for both microbiology and macroecology. Here, we identify and review important conceptual, computational, and methodological challenges and opportunities in microbial macroecology. Specifically, we discuss the challenges of handling and analyzing large amounts of microbiome data to understand taxa distribution and co-occurrence patterns. We also discuss approaches for modeling microbial communities based on environmental data, including information on biological interactions to make full use of available Big Data. Finally, we summarize the methods presented in a general approach aimed to aid microbiologists in addressing fundamental questions in microbial macroecology, including classical propositions (such as “everything is everywhere, but the environment selects”) as well as applied ecological problems, such as those posed by human induced global environmental changes.