There is a long history of archaeologists and forensic scientists using pollen found in a dust sample to identify its geographic origin or history. Such palynological approaches have important limitations as they require time-consuming identification of pollen grains, a priori knowledge of plant species distributions, and a sufficient diversity of pollen types to permit spatial or temporal identification. We demonstrate an alternative approach based on DNA sequencing analyses of the fungal diversity found in dust samples. Using nearly 1,000 dust samples collected from across the continental U.S., our analyses identify up to 40,000 fungal taxa from these samples, many of which exhibit a high degree of geographic endemism. We develop a statistical learning algorithm via discriminant analysis that exploits this geographic endemicity in the fungal diversity to correctly identify samples to within a few hundred kilometers of their geographic origin with high probability. In addition, our statistical approach provides a measure of certainty for each prediction, in contrast with current palynology methods that are almost always based on expert opinion and devoid of statistical inference. Fungal taxa found in dust samples can therefore be used to identify the origin of that dust and, more importantly, we can quantify our degree of certainty that a sample originated in a particular place. This work opens up a new approach to forensic biology that could be used by scientists to identify the origin of dust or soil samples found on objects, clothing, or archaeological artifacts.

Recent advances in bioinformatics have made high-throughput microbiome data widely available, and new statistical tools are required to maximize the information gained from these data. For example, analysis of high-dimensional microbiome data from designed experiments remains an open area in microbiome research. Contemporary analyses work on metrics that summarize collective properties of the microbiome, but such reductions preclude inference on the fine-scale effects of environmental stimuli on individual microbial taxa. Other approaches model the proportions or counts of individual taxa as response variables in mixed models, but these methods fail to account for complex correlation patterns among microbial communities. In this paper, we propose a novel Bayesian mixed-effects model that exploits cross-taxa correlations within the microbiome, a model we call MIMIX (MIcrobiome MIXed model). MIMIX offers global tests for treatment effects, local tests and estimation of treatment effects on individual taxa, quantification of the relative contribution from heterogeneous sources to microbiome variability, and identification of latent ecological subcommunities in the microbiome. MIMIX is tailored to large microbiome experiments using a combination of Bayesian factor analysis to efficiently represent dependence between taxa and Bayesian variable selection methods to achieve sparsity. We demonstrate the model using a simulation experiment and on a 2x2 factorial experiment of the effects of nutrient supplement and herbivore exclusion on the foliar fungal microbiome of Andropogon gerardii, a perennial bunchgrass, as part of the global Nutrient Network research initiative.

Summary An important problem in modern forensic analyses is identifying the provenance of materials at a crime scene, such as biological material on a piece of clothing. This procedure, which is known as geolocation, is conventionally guided by expert knowledge of the biological evidence and therefore tends to be application specific, labour intensive and often subjective. Purely data-driven methods have yet to be fully realized in this domain, because in part of the lack of a sufficiently rich source of data. However, high throughput sequencing technologies can identify tens of thousands of fungi and bacteria taxa by using DNA recovered from a single swab collected from nearly any object or surface. This microbial community, or microbiome, may be highly informative of the provenance of the sample, but data on the spatial variation of microbiomes are sparse and high dimensional and have a complex dependence structure that render them difficult to model with standard statistical tools. Deep learning algorithms have generated a tremendous amount of interest within the machine learning community for their predictive performance in high dimensional problems. We present DeepSpace: a new algorithm for geolocation that aggregates over an ensemble of deep neural network classifiers trained on randomly generated Voronoi partitions of a spatial domain. The DeepSpace algorithm makes remarkably good point predictions; for example, when applied to the microbiomes of over 1300 dust samples collected across continental USA, more than half of geolocation predictions produced by this model fall less than 100 km from their true origin, which is a 60% reduction in error from competing geolocation methods. Moreover, we apply DeepSpace to a novel data set of global dust samples collected from nearly 30 countries, finding that dust-associated fungi alone predict a sample's country of origin with nearly 90% accuracy.

No abstract

The United States Environmental Protection Agency has established a large network of stations to monitor fine particulate matter of <2.5 µm (PM2.5) that is known to be harmful to human health. Unfortunately, the network has limited spatial coverage, and stations often only measure PM2.5 every few days. Satellite‐measured aerosol optical depth (AOD) is a low‐cost surrogate with greater spatiotemporal coverage, and spatial regression models have established that including AOD as a covariate improves the spatial interpolation of PM2.5. However, AOD is often missing, and our analysis reveals that the conditions that lead to missing AOD are also conducive to high AOD. Therefore, naïve interpolation that ignores informative missingness may lead to bias. We propose a joint hierarchical model for PM2.5 and AOD that accounts for informatively missing AOD. We conduct a simulation study of the effects of ignoring informative missingness in the covariate and evaluate the performance of the proposed model. We apply the method to map daily PM2.5 in the Southeastern United States. Our analysis reveals statistically significant informative missingness and relationships between PM2.5 and AOD in many seasons after accounting for meteorological and land‐use variables.