Resolving Non-identifiability Mitigates Bias in Models of Neural Tuning and Functional Coupling.

In the brain, all neurons are driven by the activity of other neurons, some of which maybe simultaneously recorded, but most are not. As such, models of neuronal activity need to account for simultaneously recorded neurons and the influences of unmeasured neurons. This can be done through inclusion of model terms for observed external variables (e.g., tuning to stimuli) as well as terms for latent sources of variability. Determining the influence of groups of neurons on each other relative to other influences is important to understand brain functioning. The parameters of statistical models fit to data are commonly used to gain insight into the relative importance of those influences. Scientific interpretation of models hinge upon unbiased parameter estimates. However, evaluation of biased inference is rarely performed and sources of bias are poorly understood. Through extensive numerical study and analytic calculation, we show that common inference procedures and models are typically biased. We demonstrate that accurate parameter selection before estimation resolves model non-identifiability and mitigates bias. In diverse neurophysiology data sets, we found that contributions of coupling to other neurons are often overestimated while tuning to exogenous variables are underestimated in common methods. We explain heterogeneity in observed biases across data sets in terms of data statistics. Finally, counter to common intuition, we found that model non-identifiability contributes to bias, not variance, making it a particularly insidious form of statistical error. Together, our results identify the causes of statistical biases in common models of neural data, provide inference procedures to mitigate that bias, and reveal and explain the impact of those biases in diverse neural data sets.

The Functional Biogeography of eDNA Metacommunities in the Post-Fire Landscape of the Angeles National Forest.

Wildfires have continued to increase in frequency and severity in Southern California due in part to climate change. To gain a further understanding of microbial soil communities' response to fire and functions that may enhance post-wildfire resilience, soil fungal and bacterial microbiomes were studied from different wildfire areas in the Gold Creek Preserve within the Angeles National Forest using 16S, FITS, 18S, 12S, PITS, and COI amplicon sequencing. Sequencing datasets from December 2020 and June 2021 samplings were analyzed using QIIME2, ranacapa, stats, vcd, EZBioCloud, and mixomics. Significant differences were found among bacterial and fungal taxa associated with different fire areas in the Gold Creek Preserve. There was evidence of seasonal shifts in the alpha diversity of the bacterial communities. In the sparse partial least squares analysis, there were strong associations (r > 0.8) between longitude, elevation, and a defined cluster of Amplicon Sequence Variants (ASVs). The Chi-square test revealed differences in fungi−bacteria (F:B) proportions between different trails (p = 2 × 10−16). sPLS results focused on a cluster of Green Trail samples with high elevation and longitude. Analysis revealed the cluster included the post-fire pioneer fungi Pyronema and Tremella. Chlorellales algae and possibly pathogenic Fusarium sequences were elevated. Bacterivorous Corallococcus, which secretes antimicrobials, and bacterivorous flagellate Spumella were associated with the cluster. There was functional redundancy in clusters that were differently composed but shared similar ecological functions. These results implied a set of traits for post-fire resiliency. These included photo-autotrophy, mineralization of pyrolyzed organic matter and aromatic/oily compounds, potential pathogenicity and parasitism, antimicrobials, and N-metabolism.

Clinical, social, and policy factors in COVID-19 cases and deaths: methodological considerations for feature selection and modeling in county-level analyses.

BACKGROUND: There is a need to evaluate how the choice of time interval contributes to the lack of consistency of SDoH variables that appear as important to COVID-19 disease burden within an analysis for both case counts and death counts. METHODS: This study identified SDoH variables associated with U.S county-level COVID-19 cumulative case and death incidence for six different periods: the first 30, 60, 90, 120, 150, and 180 days since each county had COVID-19 one case per 10,000 residents. The set of SDoH variables were in the following domains: resource deprivation, access to care/health resources, population characteristics, traveling behavior, vulnerable populations, and health status. A generalized variance inflation factor (GVIF) analysis was used to identify variables with high multicollinearity. For each dependent variable, a separate model was built for each of the time periods. We used a mixed-effect generalized linear modeling of counts normalized per 100,000 population using negative binomial regression. We performed a Kolmogorov-Smirnov goodness of fit test, an outlier test, and a dispersion test for each model. Sensitivity analysis included altering the county start date to the day each county reached 10 COVID-19 cases per 10,000. RESULTS: Ninety-seven percent (3059/3140) of the counties were represented in the final analysis. Six features proved important for both the main and sensitivity analysis: adults-with-college-degree, days-sheltering-in-place-at-start, prior-seven-day-median-time-home, percent-black, percent-foreign-born, over-65-years-of-age, black-white-segregation, and days-since-pandemic-start. These variables belonged to the following categories: COVID-19 related, vulnerable populations, and population characteristics. Our diagnostic results show that across our outcomes, the models of the shorter time periods (30 days, 60 days, and 900 days) have a better fit. CONCLUSION: Our findings demonstrate that the set of SDoH features that are significant for COVID-19 outcomes varies based on the time from the start date of the pandemic and when COVID-19 was present in a county. These results could assist researchers with variable selection and inform decision makers when creating public health policy.

Prediction of Spread and Regional Development of Hop Powdery Mildew: A Network Analysis.

Dispersal is a fundamental aspect of epidemic development at multiple spatial scales, including those that extend beyond the borders of individual fields and to the landscape level. In this research, we used the powdery mildew of the hop pathosystem (caused by Podosphaera macularis) to formulate a model of pathogen dispersal during spring (May to June) and early summer (June to July) at the intermediate scale between synoptic weather systems and microclimate (mesoscale) based on a census of commercial hop yards during 2014 to 2017 in a production region in western Oregon. This pathosystem is characterized by a low level of overwintering of the pathogen as a result of absence of the ascigerious stage of the fungus and consequent annual cycles of localized survival via bud perennation and pathogen spread by windborne dispersal. An individual hop yard was considered a node in the model, whose disease status in a given month was expressed as a nonlinear function of disease incidence in the preceding month, susceptibility to two races of the fungus, and disease spread from other nodes as influenced by their disease incidence, area, distance away, and wind run and direction in the preceding month. Parameters were estimated by maximum likelihood over all 4 years but were allowed to vary for time transition periods from May to June and from June to July. The model accounted for 34 to 90% of the observed variation in disease incidence at the field level, depending on the year and season. Network graphs and analyses suggest that dispersal was dominated by relatively localized dispersal events (<2 km) among the network of fields, being mostly restricted to the same or adjacent farms. When formed, predicted disease attributable to dispersal from other hop yards (edges) associated with longer distance dispersal was more frequent in the June to July time transition. Edges with a high probability of disease transmission were formed in instances where yards were in close proximity or where disease incidence was relatively high in large hop yards, as moderated by wind run. The modeling approach provides a flexible and generalizable framework for understanding and predicting pathogen dispersal at the regional level as well as the implications of network connectivity on epidemic development.

Sparse, Predictive, and Interpretable Functional Connectomics with UoILasso.

Network formation from neural activity is a foundational problem in systems neuroscience. Functional networks, after downstream analysis, can provide key insights into the nature of neurobiological structure and computation. The validity of such insights hinges on accurate selection and estimation of the edges connecting nodes. However, commonly used statistical inference procedures generally fail to identify the correct features, and further introduce consequential bias in the estimates. To address these issues, we developed Union of Intersections (UoI), a flexible, modular, and scalable framework for enhanced statistical feature selection and estimation. Methods based on UoI perform feature selection and feature estimation through intersection and union operations, respectively. In the context of linear regression (specifically UoILasso), we summarize extensive numerical investigation on synthetic data to demonstrate tight control of false-positives and false-negatives in feature selection with low-bias and low-variance estimates of selected parameters, while maintaining high-quality prediction accuracy. We demonstrate, with UoILasso, the extraction of sparse, predictive, and interpretable functional networks from human electrocorticography recordings during speech production and the inference of parsimonious coupling models from nonhuman primate single-unit recordings during reaching tasks. Our results establish that UoILasso generates interpretable and predictive functional connectivity networks.