Dataset Information

Modeling spatiotemporal abundance of mobile wildlife in highly variable environments using boosted GAMLSS hurdle models.

ABSTRACT: Modeling organism distributions from survey data involves numerous statistical challenges, including accounting for zero-inflation, overdispersion, and selection and incorporation of environmental covariates. In environments with high spatial and temporal variability, addressing these challenges often requires numerous assumptions regarding organism distributions and their relationships to biophysical features. These assumptions may limit the resolution or accuracy of predictions resulting from survey-based distribution models. We propose an iterative modeling approach that incorporates a negative binomial hurdle, followed by modeling of the relationship of organism distribution and abundance to environmental covariates using generalized additive models (GAM) and generalized additive models for location, scale, and shape (GAMLSS). Our approach accounts for key features of survey data by separating binary (presence-absence) from count (abundance) data, separately modeling the mean and dispersion of count data, and incorporating selection of appropriate covariates and response functions from a suite of potential covariates while avoiding overfitting. We apply our modeling approach to surveys of sea duck abundance and distribution in Nantucket Sound (Massachusetts, USA), which has been proposed as a location for offshore wind energy development. Our model results highlight the importance of spatiotemporal variation in this system, as well as identifying key habitat features including distance to shore, sediment grain size, and seafloor topographic variation. Our work provides a powerful, flexible, and highly repeatable modeling framework with minimal assumptions that can be broadly applied to the modeling of survey data with high spatiotemporal variability. Applying GAMLSS models to the count portion of survey data allows us to incorporate potential overdispersion, which can dramatically affect model results in highly dynamic systems. Our approach is particularly relevant to systems in which little a priori knowledge is available regarding relationships between organism distributions and biophysical features, since it incorporates simultaneous selection of covariates and their functional relationships with organism responses.

SUBMITTER: Smith A

PROVIDER: S-EPMC6405508 | biostudies-literature | 2019 Mar

REPOSITORIES: biostudies-literature

ACCESS DATA

Publications

Modeling spatiotemporal abundance of mobile wildlife in highly variable environments using boosted GAMLSS hurdle models.

Smith Adam A Hofner Benjamin B Lamb Juliet S JS Osenkowski Jason J Allison Taber T Sadoti Giancarlo G McWilliams Scott R SR Paton Peter P

Ecology and evolution 20190214 5

Modeling organism distributions from survey data involves numerous statistical challenges, including accounting for zero-inflation, overdispersion, and selection and incorporation of environmental covariates. In environments with high spatial and temporal variability, addressing these challenges often requires numerous assumptions regarding organism distributions and their relationships to biophysical features. These assumptions may limit the resolution or accuracy of predictions resulting from ...[more]

PMID: 30891185

Similar Datasets

Project description:MotivationThe recently proposed knockoff filter is a general framework for controlling the false discovery rate (FDR) when performing variable selection. This powerful new approach generates a 'knockoff' of each variable tested for exact FDR control. Imitation variables that mimic the correlation structure found within the original variables serve as negative controls for statistical inference. Current applications of knockoff methods use linear regression models and conduct variable selection only for variables existing in model functions. Here, we extend the use of knockoffs for machine learning with boosted trees, which are successful and widely used in problems where no prior knowledge of model function is required. However, currently available importance scores in tree models are insufficient for variable selection with FDR control.ResultsWe propose a novel strategy for conducting variable selection without prior model topology knowledge using the knockoff method with boosted tree models. We extend the current knockoff method to model-free variable selection through the use of tree-based models. Additionally, we propose and evaluate two new sampling methods for generating knockoffs, namely the sparse covariance and principal component knockoff methods. We test and compare these methods with the original knockoff method regarding their ability to control type I errors and power. In simulation tests, we compare the properties and performance of importance test statistics of tree models. The results include different combinations of knockoffs and importance test statistics. We consider scenarios that include main-effect, interaction, exponential and second-order models while assuming the true model structures are unknown. We apply our algorithm for tumor purity estimation and tumor classification using Cancer Genome Atlas (TCGA) gene expression data. Our results show improved discrimination between difficult-to-discriminate cancer types.Availability and implementationThe proposed algorithm is included in the KOBT package, which is available at https://cran.r-project.org/web/packages/KOBT/index.html.Supplementary informationSupplementary data are available at Bioinformatics online.

Project description:Many human activities contaminate terrestrial and aquatic environments with numerous chemical pollutants that not only directly alter the environment but also affect microbial communities in ways that are potentially concerning to human health, such as selecting for the spread of antibiotic-resistance genes (ARGs) through horizontal gene transfer. In the present study, metagenomes available in the public domain from polluted (with antibiotics, with petroleum, with metal mining, or with coal-mining effluents) and unpolluted terrestrial and aquatic environments were compared to examine whether pollution has influenced the abundance and composition of ARGs and mobile elements, with specific focus on IS26 and class 1 integrons (intI1). When aggregated together, polluted environments had a greater relative abundance of ARGs than unpolluted environments and a greater relative abundance of IS26 and intI1. In general, chemical pollution, notably with petroleum, was associated with an increase in the prevalence of ARGs linked to multidrug efflux pumps. Included in the suite of efflux pumps were mexK, mexB, mexF, and mexW that are polyspecific and whose substrate ranges include multiple classes of critically important antibiotics. Also, in some instances, β-lactam resistance (TEM181 and OXA-541) genes increased, and genes associated with rifampicin resistance (RNA polymerases subunits rpoB and rpoB2) decreased in relative abundance. This meta-analysis suggests that different types of chemical pollution can enrich populations that carry efflux pump systems associated with resistance to multiple classes of medically critical antibiotics.IMPORTANCEThe United Nations has identified chemical pollution as being one of the three greatest threats to environmental health, through which the evolution of antimicrobial resistance, a seminally important public health challenge, may be favored. While this is a very plausible outcome of continued chemical pollution, there is little evidence or research evaluating this risk. The objective of the present study was to examine existing metagenomes from chemically polluted environments and evaluate whether there is evidence that pollution increases the relative abundance of genes and mobile genetic elements that are associated with antibiotic resistance. The key finding is that for some types of pollution, particularly in environments exposed to petroleum, efflux pumps are enriched, and these efflux pumps can confer resistance to multiple classes of medically important antibiotics that are typically associated with Pseudomonas spp. or other Gram-negative bacteria. This finding makes clear the need for more investigation on the impact of chemical pollution on the environmental reservoir of ARGs and their association with mobile genetic elements that can contribute to horizontal gene transfer events.

Dataset Information

Modeling spatiotemporal abundance of mobile wildlife in highly variable environments using boosted GAMLSS hurdle models.

Publications

Modeling spatiotemporal abundance of mobile wildlife in highly variable environments using boosted GAMLSS hurdle models.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets