Single Covariate Research Articles

Multi-covariate non-stationary flood frequency analysis for assessing flood hazard over the Mahanadi River basin, India

ABSTRACT Due to climate change and anthropogenic activities, the river’s peak flow is changing, which prompts exploring the non-stationary flood frequency analysis. Most of the past studies modelled the non-stationarity of floods, considering time and/or a single dominant covariate influencing the flood flows, but they are inadequate. This study proposes a multi-covariate non-stationary flood frequency model (MC-NSFFM) and investigates the influence of different covariates, such as rainfall, modified reservoir index (MRI), and climate indices, for the Mahanadi River basin in India. The models’ performances are evaluated by comparing them with the results of stationary and single-covariate non-stationary flood frequency models (SC-NSFFMs) using standard performance measures and return period analysis. The study showed that the MC-NSFFMs performed better than stationary and SC-NSFFMs. The flood return levels corresponding to the MC-NSFFMs are 4% to 22% higher than the stationary models, indicating the rise in flood risks due to climate change and anthropogenic activities.

Non-stationarity of runoff and sediment load and its drivers under climate change and anthropogenic activities in Dongting Lake Basin

Analysing non-stationarity in runoff and sediment load is crucial for effective water resource management in the Dongting Lake basin amid climate change and human impacts. Using the Mann-Kendall test, Generalized Additive Models for Location, Scale, and Shape framework, and Random Forest models, we evaluated non-stationarity and its drivers in the annual runoff and annual sediment load series at eight hydrological stations from 1961 to 2021. These stations include three inflow sites at the Jingjiang Three Outlets (Ouchi, Songzi, and Hudu Rivers), four inflow sites in the Four Rivers basin (Xiang, Zi, Yuan, and Li Rivers), and one outflow site at Chenglingji. Results revealed a significant decrease in annual runoff at the Three Outlets and Chenglingji, while the Four Rivers basin showed no significant trend. The non-stationary models with multiple physically-based covariates better captured non-stationarity compared to single covariate models. Annual rainfall was a key contributor to annual runoff in the Four Rivers basin, while reservoir storage capacity played a more dominant role in the Three Outlets. At Chenglingji station, both factors significantly influenced annual runoff. For annual sediment load, reservoir storage capacity emerged as the most critical factor across all regions. These findings provide a basis for improving runoff and sediment regulation in the Dongting Lake basin.

Integrating heterogeneous information for modeling non-stationarity of extreme precipitation in the Yangtze River Basin

Extreme precipitation exhibits non-stationary characteristics due to climate change and land use/cover change in recent years. In the literature, probability distributions with varying parameters associated with covariates have been applied to describe the non-stationarity of extreme precipitation. However, there is no consensus on the optimal distributions and the number and type of covariates for constructing non-stationary models. This study comprehensively investigates and determines the optimal distribution models, the type of covariates, and the number of covariates to best characterize the non-stationarity of extreme precipitation, using the Yangtze River Basin as the study area. The time series of four indices representing extreme precipitation of different durations and intensities are fitted by stationary and non-stationary Generalized Additive Models for Location, Scale, and Shape. The models are constructed using seven distributions and four types (time, global climate, local climate, and land cover) of covariates. Each model incorporates one to three covariates simultaneously. The results show that the most preferred distribution is the Generalized Extreme Value distribution for very wet-day precipitation, while the Log-Normal distribution is more suitable for other extreme precipitation indices. For the type of covariates, when using a single covariate, the proportion of stations using a global climate covariate as the best covariate (22.3 %–33.2 % of 193 stations) is close to that using a local climate covariate as the best covariate (22.3 %–27.5 %). The proportion of stations using land cover as the best covariate is much lower (7.3 %–10.9 %), while that using time as the best covariate is the lowest (1.6 %–4.1 %). When using multiple covariates, the models with both global and local climate covariates are mostly selected (8.3 % to 17.6 %) for all extreme precipitation indices. For the number of covariates, one or two covariates are enough to describe the non-stationary behaviors of all extreme precipitation indices. This study emphasizes the necessity to use non-stationary models for extreme precipitation frequency analysis with careful consideration and selection of distributions and covariates.

Synthesis estimators for transportability with positivity violations by a continuous covariate

Abstract Studies intended to estimate the effect of a treatment, like randomized trials, may not be sampled from the desired target population. To correct for this discrepancy, estimates can be transported to the target population. Methods for transporting between populations are often premised on a positivity assumption, such that all relevant covariate patterns in one population are also present in the other. However, eligibility criteria, particularly in the case of trials, can result in violations of positivity when transporting to external populations. To address nonpositivity, a synthesis of statistical and mathematical models can be considered. This approach integrates multiple data sources (e.g. trials, observational, pharmacokinetic studies) to estimate treatment effects, leveraging mathematical models to handle positivity violations. This approach was previously demonstrated for positivity violations by a single binary covariate. Here, we extend the synthesis approach for positivity violations with a continuous covariate. For estimation, two novel augmented inverse probability weighting estimators are proposed. Both estimators are contrasted with other common approaches for addressing nonpositivity. Empirical performance is compared via Monte Carlo simulation. Finally, the competing approaches are illustrated with an example in the context of two-drug vs. one-drug antiretroviral therapy on CD4 T cell counts among women with HIV.

Measurement and Uncertainty Preserving Parametric Modeling for Continuous Latent Variables With Discrete Indicators and External Variables

Research in education and behavioral sciences often involves the use of latent variable models that are related to indicators, as well as related to covariates or outcomes. Such models are subject to interpretational confounding, which occurs when fitting the model with covariates or outcomes alters the results for the measurement model. This has received attention in models for continuous observable variables but to date has not been examined in the context of discrete variables. This work demonstrates that interpretational confounding can occur in models for discrete variables, and develops a multistage Bayesian estimation approach to deal with this problem. The key features of this approach are that it is (a) measurement preserving, in that it precludes the possibility of interpretational confounding, and (b) uncertainty preserving, in that the uncertainty from the earlier stage of estimating the measurement model is propagated to the second stage of estimating the relations between the latent variable(s) and any covariates or outcomes. Previous work on these methods had only considered models for continuous observed variables, and software was limited to models with a single latent variable and either covariates or outcomes. This work extends the approach and software to a more general class of solutions, including discrete variables, illustrating the procedures with analyses of real data. Functions for conducting the analyses in widely available software are provided.

A censored quantile transformation model for Alzheimer’s Disease data with multiple functional covariates

Abstract Alzheimer’s disease (AD) is a progressive disease that starts from mild cognitive impairment and may eventually lead to irreversible memory loss. It is imperative to explore the risk factors associated with the conversion time to AD that is usually right-censored. Classical statistical models like mean regression and Cox models fail to quantify the impact of risk factors across different quantiles of a response distribution, and previous research has primarily focused on modelling a single functional covariate, possibly overlooking the interdependence among multiple functional covariates and other crucial features of the distribution. To address these issues, this paper proposes a multivariate functional censored quantile regression model based on dynamic power transformations, which relaxes the global linear assumption and provides more robustness and flexibility. Uniform consistency and weak convergence of the quantile process are established. Simulation studies suggest that the proposed method outperforms the existing approaches. Real data analysis shows the importance of both left and right hippocampal radial distance curves for predicting the conversion time to AD at different quantile levels.

Comparison of brace to observation in stable, radiological developmental dysplasia of the hip: a protocol for a global multicentre non-inferiority randomised trial

IntroductionBrace treatment is common to address radiological dysplasia in infants with developmental dysplasia of the hip (DDH); however, it is unclear whether bracing provides significant benefit above careful observation by...

Do Common Risk Adjustment Methods Do Their Job Well if Center Effects are Correlated With the Center-Specific Mean Values of Patient Characteristics?

Direct and indirect standardization are well-established approaches to performing risk adjustment when comparing outcomes between healthcare providers. However, it is an open question whether they work well when there is an association between the center effects and the distributions of the patient characteristics in these centers. We try to shed further light on the impact of such an association. We construct an artificial case study with a single covariate, in which centers can be classified as performing above, on, or below average, and the center effects correlate with center-specific mean values of a patient characteristic, as a consequence of differential quality improvement. Based on this case study, direct standardization and indirect standardization-based on marginal as well as conditional models-are compared with respect to systematic differences between their results. Systematic differences between the methods were observed. All methods produced results that partially reflect differences in mean age across the centers. This may mask the classification as above, on, or below average. The differences could be explained by an inspection of the parameter estimates in the models fitted. In case of correlations of center effects with center-specific mean values of a covariate, different risk adjustment methods can produce systematically differing results. This suggests the routine use of sensitivity analyses. Center effects in a conditional model need not reflect the position of a center above or below average, questioning its use in defining the truth. Further empirical investigations are necessary to judge the practical relevance of these findings.

Age and Sex Differences in the Genetic Architecture of Measures of Subjective Health: Relationships With Physical Health, Depressive Symptoms, and Episodic Memory.

Subjective health (SH) is not just an indicator of physical health, but also reflects active cognitive processing of information about one's own health and has been associated with emotional health measures, such as neuroticism and depression. Behavior genetic approaches investigate the genetic architecture of SH, that is, genetic and environmental influences on individual differences in SH and associations with potential components such as physical, cognitive, and emotional health. Previous twin analyses have been limited by sex, sample size, age range, and focus on single covariates. The current analysis used data from 24,173 adults ranging in age from 40 to 90 years from the international Interplay of Genes and Environment across Multiple Studies consortium to investigate the genetic architecture of 3 measures of SH: self-rated health, health compared to others, and impact of health on activities. Independent pathways model of SH included physical health, depressive symptoms, and episodic memory, with age, sex, and country included as covariates. Most or all of the genetic variance for SH measures were shared with physical health, depressive symptoms, and episodic memory. Genetic architecture of SH differed across measures, age groups (40-65, 66-90), and sexes. Age comparisons indicated stronger correlations with all 3 covariates in older adults, often resulting from greater shared genetic variance. The predictive value of SH has been amply demonstrated. The higher genetic contributions to associations between SH and its components in older adults support the increasing conceptualization with age of SH as an intuitive summation of one's vital reserve.

Quantifying the Impact of Alternative Definitions of Sepsis-Associated Acute Kidney Injury on its Incidence and Outcomes: A Systematic Review and Meta-Analysis.

To derive a pooled estimate of the incidence and outcomes of sepsis-associated acute kidney injury (SA-AKI) in ICU patients and to explore the impact of differing definitions of SA-AKI on these estimates. Medline, Medline Epub, EMBASE, and Cochrane CENTRAL between 1990 and 2023. Randomized clinical trials and prospective cohort studies of adults admitted to the ICU with either sepsis and/or SA-AKI. Data were extracted in duplicate. Risk of bias was assessed using adapted standard tools. Data were pooled using a random-effects model. Heterogeneity was assessed by using a single covariate logistic regression model. The primary outcome was the proportion of participants in ICU with sepsis who developed AKI. A total of 189 studies met inclusion criteria. One hundred fifty-four reported an incidence of SA-AKI, including 150,978 participants. The pooled proportion of patients who developed SA-AKI across all definitions was 0.40 (95% CI, 0.37-0.42) and 0.52 (95% CI, 0.48-0.56) when only the Risk Injury Failure Loss End-Stage, Acute Kidney Injury Network, and Improving Global Outcomes definitions were used to define SA-AKI. There was significant variation in the incidence of SA-AKI depending on the definition of AKI used and whether AKI defined by urine output criteria was included; the incidence was lowest when receipt of renal replacement therapy was used to define AKI (0.26; 95% CI, 0.24-0.28), and highest when the Acute Kidney Injury Network score was used (0.57; 95% CI, 0.45-0.69; p < 0.01). Sixty-seven studies including 29,455 participants reported at least one SA-AKI outcome. At final follow-up, the proportion of patients with SA-AKI who had died was 0.48 (95% CI, 0.43-0.53), and the proportion of surviving patients who remained on dialysis was 0.10 (95% CI, 0.04-0.17). SA-AKI is common in ICU patients with sepsis and carries a high risk of death and persisting kidney impairment. The incidence and outcomes of SA-AKI vary significantly depending on the definition of AKI used.

Potential of dynamic ocean management strategies for western Pacific leatherback sea turtle bycatch mitigation in New Zealand

Western Pacific leatherback sea turtles (Dermochelys coriacea) are a priority bycatch mitigation concern due to the projected extinction of the population before the end of the 21st century. The species regularly occurs as bycatch in gillnet and surface longline fisheries. Here, we explore the potential for dynamic ocean management in an emerging hotspot of leatherback sea turtle bycatch in the New Zealand pelagic longline fishery. We compared spatial areas of different sizes built from single oceanographic covariates as well as built from a composite risk surface developed through ensemble random forests. We found that, individually, the Okubo–Weiss parameter, sea surface temperature (SST) anomaly, SST, moon phase, and distance to the SST front were important oceanographic covariates for leatherback sea turtle bycatch. However, the spatial areas built from the composite risk surface were the most effective at discriminating sets with and without bycatch across a range of risk cutoffs. When we also considered implementation metrics of spatial area and coherence as part of performance, the area derived from the composite risk surface with a risk of interaction per set greater than 52% performed best. This spatial area was ephemeral, occurring 1 or 2 weeks each year, and localized, occurring along the north coast of East Cape in the North Island of New Zealand. The apparent presence of discrete spatial areas with elevated risk may be useful to inform future management in the area. Considering implementation metrics in defining utility was useful for identifying tradeoffs between the total size and the underlying covariates delineating a spatial area. As such, we recommend these types of metrics to be included when designing spatial bycatch mitigation strategies elsewhere.

A data-adaptive method for investigating effect heterogeneity with high-dimensional covariates in Mendelian randomization

BackgroundMendelian randomization is a popular method for causal inference with observational data that uses genetic variants as instrumental variables. Similarly to a randomized trial, a standard Mendelian randomization analysis estimates the population-averaged effect of an exposure on an outcome. Dividing the population into subgroups can reveal effect heterogeneity to inform who would most benefit from intervention on the exposure. However, as covariates are measured post-“randomization”, naive stratification typically induces collider bias in stratum-specific estimates.MethodWe extend a previously proposed stratification method (the “doubly-ranked method”) to form strata based on a single covariate, and introduce a data-adaptive random forest method to calculate stratum-specific estimates that are robust to collider bias based on a high-dimensional covariate set. We also propose measures based on the Q statistic to assess heterogeneity between stratum-specific estimates (to understand whether estimates are more variable than expected due to chance alone) and variable importance (to identify the key drivers of effect heterogeneity).ResultWe show that the effect of body mass index (BMI) on lung function is heterogeneous, depending most strongly on hip circumference and weight. While for most individuals, the predicted effect of increasing BMI on lung function is negative, it is positive for some individuals and strongly negative for others.ConclusionOur data-adaptive approach allows for the exploration of effect heterogeneity in the relationship between an exposure and an outcome within a Mendelian randomization framework. This can yield valuable insights into disease aetiology and help identify specific groups of individuals who would derive the greatest benefit from targeted interventions on the exposure.

Predicting Imminent Cyanobacterial Blooms in Lakes Using Incomplete Timely Data

AbstractToxic cyanobacterial blooms (CBs) are becoming more frequent globally, posing a threat to freshwater ecosystems. While making long‐range forecasts is overly challenging, predicting imminent CBs is possible from precise monitoring data of the underlying covariates. It is, however, infeasibly costly to conduct precise monitoring on a large scale, leaving most lakes unmonitored or only partially monitored. The challenge is hence to build a predictive model that can use the incomplete, partially‐monitored data to make near‐future CB predictions. By using 30 years of monitoring data for 78 water bodies in Alberta, Canada, combined with data of watershed characteristics (including natural land cover and anthropogenic land use) and meteorological conditions, we train a Bayesian network that predicts future 2‐week CB with an area under the curve (AUC) of 0.83. The only monitoring data that the model needs to reach this level of accuracy are whether the cell count and Secchi depth are low, medium, or high, which can be estimated by advanced high‐resolution imaging technology or trained local citizens. The model is robust against missing values as in the absence of any single covariate, it performs with an AUC of at least 0.78. While taking a major step toward reduced‐cost, less data‐intensive CB forecasting, our results identify those key covariates that are worth the monitoring investment for highly accurate predictions.

Adaptive selection of the optimal strategy to improve precision and power in randomized trials.

Benkeser etal. demonstrate how adjustment for baseline covariates in randomized trials can meaningfully improve precision for a variety of outcome types. Their findings build on a long history, starting in 1932 with R.A. Fisher and including more recent endorsements by the U.S. Food and Drug Administration and the European Medicines Agency. Here, we address an important practical consideration: how to select the adjustment approach-which variables and in which form-to maximize precision, while maintaining Type-I error control. Balzer etal. previously proposed Adaptive Pre-specification within TMLE to flexibly and automatically select, from a prespecified set, the approach that maximizes empirical efficiency in small trials (N< 40). To avoid overfitting with few randomized units, selection was previously limited to working generalized linear models, adjusting for a single covariate. Now, we tailor Adaptive Pre-specification to trials with many randomized units. Using V-fold cross-validation and the estimated influence curve-squared as the loss function, we select from an expanded set of candidates, including modern machine learning methods adjusting for multiple covariates. As assessed in simulations exploring a variety of data-generating processes, our approach maintains Type-I error control (under the null) and offers substantial gains in precision-equivalent to 20%-43% reductions in sample size for the same statistical power. When applied to real data from ACTG Study 175, we also see meaningful efficiency improvements overall and within subgroups.

Habitat occupancy of the threatened Diademed Plover (Phegornis mitchellii) is not affected by llama grazing or peatland size, but declines with peatland humidity.

Many habitat-specialist organisms occur in distinct, patchy habitat, yet do not occupy all patches, and an important question is why apparently suitable habitat remains unoccupied. We examined factors influencing patch occupancy in near-threatened, little-known Diademed Plovers (Phegornis mitchellii), arguably the bird most specialized to life in High Andean peatlands. Andean peatlands are well-suited to occupancy modelling because they are discrete patches of humid habitat within a matrix of high-altitude steppe. We hypothesized that Diademed Plovers occupy preferably larger and more humid peatlands, and avoid peatlands used for grazing by llamas and vicuñas, which may trample vegetation and nests. From December 2021 to February 2022 (breeding season), we conducted plover occupancy surveys (2-4) on 40 peatlands at Lagunas de Vilama, a landscape of arid steppe and wetlands above 4,500 m in NW Argentina. We measured peatland size, grazing pressure, topographic and remotely-sensed variables that correlate with humidity, and incorporated these as covariates in occupancy models. Occupancy models showed that more than 50% of the studied peatlands were used by Diademed Plovers and most showed signs of reproduction, highlighting the importance of the Vilama Wetlands for Diademed Plover conservation. Within peatlands, Diademed Plovers were most often associated with headwaters. The top ranked occupancy model included constant detection, random spatial effects, and a single occupancy covariate: mean NDWI (Normalized Difference Water Index, an index correlated with water content and humidity) over the previous three years. Contrary to our prediction, Diademed Plovers preferred less water-saturated peatlands (lower NDWI), possibly to avoid nest flooding. This may be especially important in wet years, like the year when we conducted our surveys. Neither peatland size nor grazing by llamas and vicuñas affected peatland use by Diademed Plovers, suggesting that llama grazing at current levels may be compatible with plover conservation. For organisms that specialize on humid habitats, such as peatlands, factors affecting occupancy may vary temporally with variation in climate, and we recommend follow-up surveys across multi-year timescales to untangle the impact of climate on animals' use of humid habitats.

Nonparametric estimation of stochastic frontier models with weak separability

We propose a robust and versatile approach to estimate the stochastic frontier model which avoids parametric assumptions. Our approach requires a single continuous covariate which monotonically influences the conditional mean of inefficiency. Subject to these conditions, the frontier and the conditional mean of inefficiency can be estimated nonparametrically. The estimator we propose uses local least squares and marginal integration making it easy to implement across statistical software. A range of Monte Carlo simulations suggests that when our main identification condition holds, our proposed estimator outperforms other proposals that currently exist. Finally, we provide an application to the study of undercounting COVID-19 cases across the United States. Whereas our method indicates significant undercounting, consistent with existing evidence, other nonparametric methods suggest far less undercounting.

Time to Treatment for Patients with Suspected Acute Promyelocytic Leukemia (APL)

Introduction: Acute promyelocytic leukemia (APL) is a rare subtype of acute myeloid leukemia characterized by dysplastic promyelocytes which cause aberrations in coagulation, including hemorrhage and disseminated intravascular coagulation (DIC), leading to a rapidly fatal course if treatment is not initiated promptly. Early administration of all-trans retinoic acid (ATRA) in patients suspected of having APL is crucial for reducing morbidity and mortality. However, APL still exhibits a high early death rate in part due to failure to rapidly initiate ATRA. Therefore, we conducted a retrospective study to investigate the time it takes for ATRA to be ordered and administered to patients with suspected APL at our tertiary care University hospital. We also examined demographic and clinical variables to identify any opportunities for quality improvement. Methods: We queried our electronic medical record (EMR) for all adult patients without a prior diagnosis of APL from 2017-2021 who received at least one dose of ATRA at Thomas Jefferson University Hospital. Our analysis included 91 patients. We collected a wide range of demographic and clinical data (Table 1). The time of admission was defined as the time the admission order was placed in the EMR, and the time of administration was defined as the time the nurse marked the medication as administered. A simple general linear model was used to evaluate univariate association between each single covariate and each log transformed outcome. Model selection by Akaike information criterion (AIC) was conducted to identify the best model according to goodness of fit criteria. The smaller the AIC value, the better the model fit. The parsimonious models were obtained using backward elimination of non-significant predictors (p-values &amp;gt; 0.05) from the best models according to goodness of fit criteria (AIC). Results: The mean time for interval A (time from hospital admission to ordering ATRA) was 284 minutes, the mean time for interval B (time from ATRA order being placed to ATRA administration) was 104 minutes, and the mean time for interval C (time from admission to ATRA administration) was 388 minutes. Further analysis of the time intervals is presented in Table 2. The results of the demographic and clinical variables (Table 1) show that the majority of patients were transferred from an outside hospital (89%) and admitted overnight (87%). Consequently, ATRA was most often ordered (76%) and administered (70%) overnight. Only 26 (29%) patients were ultimately confirmed to have APL. During the index admission, patients required a mean of 6 units of packed red blood cells and 9 units of platelets. A quarter of patients were noted to have signs or symptoms of bleeding upon admission. The results of the multiple linear regression analysis are shown in Table 2. Only signs or symptoms of bleeding were associated with a statistically significant decrease in time to ATRA. Patients with signs of symptoms of bleeding experienced, on average, 39% less time from admission to ATRA being ordered (95% CI: 0.396-0.946, p=0.030) and 31% shorter time from admission to ATRA administration compared to those without signs or symptoms of bleeding (95% CI: 0.489-0.967, p=0.034). Conclusions: Our study highlights that, even at a large, resource-rich, academic medical center, the time to ATRA administration in patients with suspected APL is longer than optimal. The presence of bleeding was the only variable that significantly decreased the time to ATRA. Earlier identification of bleeding risk (including evolving DIC or thrombocytopenia), increased recognition of possible acute leukemia, and more rapid hematology/oncology consultation could help to decrease the time to ATRA and to avoid bleeding complications before they start. Additionally, we found that most patients (89%) were transferred from an outside hospital. Although we were not able to quantify the time from outside hospital admission to transfer, we suspect that the time was substantial and that there are even greater opportunities to improve the care of these patients by increasing ATRA availability at community hospitals, enhancing the education of all front-line providers regarding the prompt recognition of coagulopathy in the setting of suspected acute leukemia, and improving transfer center acceptance documentation that emphasizes the concern for APL and appropriate next steps in evaluation and treatment.

Simulating time-to-event data under the Cox proportional hazards model: assessing the performance of the non-parametric Flexible Hazards Method

Numerous methods and approaches have been developed for generating time-to-event data from the Cox Proportional Hazards (CPH) model; however, they often require specification of a parametric distribution for the baseline hazard even though the CPH model itself makes no assumptions on the distribution of the baseline hazards. In line with the semi-parametric nature of the CPH model, a recently proposed method called the Flexible Hazards Method generates time-to-event data from a CPH model using a non-parametric baseline hazard function. While the initial results of this method are promising, it has not yet been comprehensively assessed with increasing covariates or against data generated under parametric baseline hazards. To fill this gap, we conducted a comprehensive study to benchmark the performance of the Flexible Hazards Method for generating data from a CPH model against parametric methods. Our results showed that with a single covariate and large enough assumed maximum time, the bias in the Flexible Hazards Method is 0.02 (with respect to the log hazard ratio) with a 95% confidence interval having coverage of 84.4%. This bias increases to 0.054 when there are 10 covariates under the same settings and the coverage of the 95% confidence interval decreases to 46.7%. In this paper, we explain the plausible reasons for this observed increase in bias and decrease in coverage as the number of covariates are increased, both empirically and theoretically, and provide readers and potential users of this method with some suggestions on how to best address these issues. In summary, the Flexible Hazards Method performs well when there are few covariates and the user wishes to simulate data from a non-parametric baseline hazard.

Modeling host–microbiome interactions to improve mechanistic understanding of aphid vectored plant pathogens

Insect transmission of plant pathogens involves multi-layered interactions between vectors, viruses, host plants and environmental factors. Adding to the complexity of vector–virus relationships are diverse microbial communities, which are hypothesized to influence pathogen transmission. Although vector–virus interaction research has flourished, the role played by microbes in vector competence and disease epidemiology remains unclear in many pathosystems. We therefore aimed to develop a novel ecological modeling approach to identify environmental drivers of complex vector–virus–microbiome interactions, particularly differences in the abundance of microbial symbionts within vector microbiomes and relationships between symbionts and the probability of virus acquisition. Our approach combines established molecular tools for profiling microbial communities with underutilized Bayesian hierarchical modeling and data integration techniques. We used a globally relevant aphid–virus pathosystem to develop custom vector–microbiome models that incorporate environmental covariates (e.g., temperature, landcover) and applied them to individual vector symbionts to identify the extent to which environmental factors drive changes in microbial communities that then influence virus acquisition by the host aphid. Specifically, we focus on the aphid obligate symbiont (Buchnera) and a wide-spread facultative symbiont (Serratia) as proof of concept to develop models for two major vector species that include a single environmental covariate (i.e., temperature). Overall, we demonstrate how community-level modeling and microbiome profiling can identify candidate microbes and environmental variables associated with vector competence. Our modeling framework can accommodate a range of microbial symbionts with different abundances, overcome spatial misalignment of data streams, and is robust to varying levels of disease incidence. Results show Buchnera relative abundance is strongly negatively associated with virus acquisition by the vector S.avenae, but not R. padi. Although Serratia was not associated with R. padi vector competence, relative abundance was influenced by differences in spring temperatures. This work lays the foundation for developing a broader modeling framework for predicting disease dynamics in agroecosystems and deploying microbiome-targeted pest management tactics.

Go beyond the limits of genetic algorithm in daily covariate selection practice

Covariate identification is an important step in the development of a population pharmacokinetic/pharmacodynamic model. Among the different available approaches, the stepwise covariate model (SCM) is the most used. However, SCM is based on a local search strategy, in which the model-building process iteratively tests the addition or elimination of a single covariate at a time given all the others. This introduces a heuristic to limit the searching space and then the computational complexity, but, at the same time, can lead to a suboptimal solution. The application of genetic algorithms (GAs) for covariate selection has been proposed as a possible solution to overcome these limitations. However, their actual use during model building is limited by the extremely high computational costs and convergence issues, both related to the number of models being tested. In this paper, we proposed a new GA for covariate selection to address these challenges. The GA was first developed on a simulated case study where the heuristics introduced to overcome the limitations affecting currently available GA approaches resulted able to limit the selection of redundant covariates, increase replicability of results and reduce convergence times. Then, we tested the proposed GA on a real-world problem related to remifentanil. It obtained good results both in terms of selected covariates and fitness optimization, outperforming the SCM.