Explore the vast range of titles available.

Many contemporary large-scale applications involve building interpretable models linking a large set of potential covariates to a response in a non-linear fashion, such as when the response is binary. Although this modelling problem has been extensively studied, it remains unclear how to control the fraction of false discoveries effectively even in high dimensional logistic regression, not to mention general high dimensional non-linear models. To address such a practical problem, we propose a new framework of ‘model-X’ knockoffs, which reads from a different perspective the knockoff procedure that was originally designed for controlling the false discovery rate in linear models. Whereas the knockoffs procedure is constrained to homoscedastic linear models with n ⩾ p, the key innovation here is that model-X knockoffs provide valid inference from finite samples in settings in which the conditional distribution of the response is arbitrary and completely unknown. Furthermore, this holds no matter the number of covariates. Correct inference in such a broad setting is achieved by constructing knockoff variables probabilistically instead of geometrically. To do this, our approach requires that the covariates are random (independent and identically distributed rows) with a distribution that is known, although we provide preliminary experimental evidence that our procedure is robust to unknown or estimated distributions. To our knowledge, no other procedure solves the controlled variable selection problem in such generality but, in the restricted settings where competitors exist, we demonstrate the superior power of knockoffs through simulations. Finally, we apply our procedure to data from a case–control study of Crohn’s disease in the UK, making twice as many discoveries as the original analysis of the same data.

Where Medicine Went Wrong explores how the idea of an average value has been misapplied to medical phenomena, distorted understanding and lead to flawed medical decisions. Through new insights into the science of complexity, traditional physiology is replaced with fractal physiology, in which variability is more indicative of health than is an average. The capricious nature of physiological systems is made conceptually manageable by smoothing over fluctuations and thinking in terms of averages. But these variations in such aspects as heart rate, breathing and walking are much more susceptible to the early influence of disease than are averages.

Rapid changes in thermal environments are threatening many species worldwide. Thermal acclimatisation may partially buffer species from the impacts of these changes, but currently, the knowledge about the temporal dynamics of acclimatisation remains limited. Moreover, acclimatisation phenotypes are typically determined in laboratory conditions that lack the variability and stochasticity that characterise the natural environment. Through a distributed lag non‐linear model (DLNM), we use field data to assess how the timing and magnitude of past thermal exposures influence thermal tolerance. We apply the model to two Scottish freshwater Ephemeroptera species living in natural thermal conditions. Model results provide evidence that rapid heat hardening effects are dramatic and reflect high rates of change in temperatures experienced over recent hours to days. In contrast, temperature change magnitude impacted acclimatisation over the course of weeks but had no impact on short‐term responses. Our results also indicate that individuals may de‐acclimatise their heat tolerance in response to cooler environments. Based on the novel insights provided by this powerful modelling approach, we recommend its wider uptake among thermal physiologists to facilitate more nuanced insights in natural contexts, with the additional benefit of providing evidence needed to improve the design of laboratory experiments.

In landscapes dominated by agriculture, conspicuous edges often occur between landscape elements. However, there is disagreement about the existence and intensity of edge effects, and information about species-specific responses remains scarce. Studying such edge effects can help elucidate functional landscape connectivity and contribute to agricultural management. We, therefore, assessed whether sun-grown coffee represents a barrier to dung beetles in an Andean agricultural landscape. We also evaluated whether the response to edge effects differs among species. We found that diversity and abundance tend to decrease from forest to sun-grown coffee and that there are sharp increases in species turnover at the forest–coffee edge. We detected several different species-specific responses to the forest–coffee edge, suggesting differences in the mobility of the species (or spillover) and in the degree of penetration that takes place from forest patches to sun-grown coffee plantations. This study demonstrates that the sun-grown coffee matrix constitutes a barrier to forest species and suggests that the forest–coffee ecotone is more complex than expected. Our results support the notion that the conservation value of native forest patches in agricultural scenarios depends on the functional connectivity of forest units in the landscape to maximize the opportunities species have to disperse through the agricultural matrix.

Herein, state‐of‐the‐art multiscale modeling methods have been described. This research includes notable molecular, micro‐, meso‐, and macroscale models for hard (polymer, metal, yarn, fiber, fiber‐reinforced polymer, and polymer matrix composites) and soft (biological tissues such as brain white matter [BWM]) composite materials. These numerical models vary from molecular dynamics simulations to finite‐element (FE) analyses and machine learning/deep learning surrogate models. Constitutive material models are summarized, such as viscoelastic hyperelastic, and emerging models like fractional viscoelastic. Key challenges such as meshing, data variability and material nonlinearity‐driven uncertainty, limitations in terms of computational resources availability, model fidelity, and repeatability are outlined with state‐of‐the‐art models. Latest advancements in FE modeling involving meshless methods, hybrid ML and FE models, and nonlinear constitutive material (linear and nonlinear) models aim to provide readers with a clear outlook on futuristic trends in composite multiscale modeling research and development. The data‐driven models presented here are of varying length and time scales, developed using advanced mathematical, numerical, and huge volumes of experimental results as data for digital models. An in‐depth discussion on data‐driven models would provide researchers with the necessary tools to build real‐time composite structure monitoring and lifecycle prediction models.

Many chemical and biological experiments involve multiple treatment factors and often it is convenient to fit a non-linear model in these factors. This non-linear model can be mechanistic, empirical or a hybrid of the two. Motivated by experiments in chemical engineering, we focus on 𝐷-optimal designs for multifactor non-linear response surfaces in general. To find and study optimal designs, we first implement conventional point and co-ordinate exchange algorithms. Next, we develop a novel multiphase optimization method to construct 𝐷-optimal designs with improved properties. The benefits of this method are demonstrated by application to two experiments involving non-linear regression models. The designs obtained are shown to be considerably more informative than designs obtained by using traditional design optimality algorithms.

Modeling plant growth using functional traits is important for understanding the mechanisms that underpin growth and for predicting new situations. We use three data sets on plant height over time and two validation methods—in‐sample model fit and leave‐one‐species‐out cross‐validation—to evaluate non‐linear growth model predictive performance based on functional traits. In‐sample measures of model fit differed substantially from out‐of‐sample model predictive performance; the best fitting models were rarely the best predictive models. Careful selection of predictor variables reduced the bias in parameter estimates, and there was no single best model across our three data sets. Testing and comparing multiple model forms is important. We developed an R package with a formula interface for straightforward fitting and validation of hierarchical, non‐linear growth models. Our intent is to encourage thorough testing of multiple growth model forms and an increased emphasis on assessing model fit relative to a model's purpose. Modeling plant growth using functional traits is important for understanding the mechanisms that underpin growth and for predicting new situations. In‐sample measures of model fit differed substantially from out‐of‐sample model predictive performance, the best fitting models were rarely the best predictive models. We developed an R package with a formula interface for straightforward fitting and validation of hierarchical, non‐linear growth models. Our intent is to encourage thorough testing of multiple growth model forms and an increased emphasis on assessing model fit relative to a model’s purpose.

Respiratory syncytial virus (RSV) is the leading cause of clinical pneumonia in children. We aimed to investigate the associations between ambient temperature, relative humidity, and pediatric RSV infections, and to assess the disease burden attributable to cold or humid conditions. Daily data on RSV hospitalizations among children aged ≤5 years, mean temperature, and relative humidity in Suzhou, China, from January 2016 to December 2019 were collected. A distributed lag nonlinear model with quasi‐Poisson regression was employed to assess the exposure‐lag‐response associations. Attributable risks were calculated to quantify the disease burden due to climatic factors. We found an inverted U‐shaped relationship between temperature and RSV infections, with the cumulative risk of RSV peaking at 7.5°C (RR = 4.30, 95% CI: 3.08–6.02). The exposure‐response curves for relative humidity exhibited a generally positive trend, peaking at 100.0% (RR = 3.14, 95% CI: 1.84–5.34). Using median values as references, the highest risk effects of extremely low (RR = 1.14, 95% CI: 1.04–1.25) and low (RR = 1.22, 95% CI: 1.12–1.32) temperatures, as well as high (RR = 1.09, 95% CI: 1.04–1.13) and extremely high (RR = 1.16, 95% CI: 1.07–1.27) relative humidity, occurred on the day of exposure and persisted for extended periods. The attributable fraction of RSV infections associated with cold or humid conditions was 55.23% (95% CI: 50.01%–64.03%) and 12.02% (95% CI: 9.36%–20.24%), respectively. The risk effect of high relative humidity was stronger in children aged 1–5 years. Our findings suggest nonlinear, lagged associations between climatic factors and pediatric RSV infections, which may inform future healthcare planning and RSV immunization strategies. Plain Language Summary Respiratory syncytial virus (RSV) is the leading pathogen responsible for clinical pneumonia in infants and young children. RSV epidemics typically occur during the winter in temperate and subtropical climates, whereas in tropical climates, peak activity is observed during the rainy season. Climatic variables play a key role in shaping RSV seasonality, and understanding these drivers is essential for enhancing epidemic forecasting, informing healthcare planning, and developing effective prevention strategies. In this study, we used a distributed lag non‐linear model to investigate the exposure‐lag‐response associations between ambient temperature and relative humidity and pediatric RSV infections in Suzhou, Eastern China. From 2016 to 2019, a total of 5,008 hospitalized cases of RSV infections among children aged ≤5 years were included in this analysis. We observed an inverted U‐shaped relationship between ambient temperature and RSV infections, while relative humidity exhibited a generally positive correlation. The risk effects associated with extreme conditions of low temperature and high relative humidity were observed on the day of exposure and extended over prolonged lag periods. Besides, cold or humid conditions could account for a substantial proportion of pediatric RSV infections. These findings will enhance early warning systems for RSV epidemics and inform healthcare planning. Key Points Temperature had an inverted U‐shape with respiratory syncytial virus infections, while relative humidity was positively correlated The risk effects of extreme climatic factors were observed on the concurrent day and persisted over extended lag periods The attributable burden of respiratory syncytial virus infections associated with cold was higher than that related to humid conditions

Hand, Foot, and Mouth Disease (HFMD) is an infectious disease that primarily affects young children. In densely populated Jiangsu Province in China, the impact of extreme meteorological factors on HFMD is a concern. We aimed to examine the association between extreme meteorological variables and HFMD infection risk using daily HFMD infections and meteorological data from 2010 to 2017 in Jiangsu Province. We used distributed lag non‐linear model (DLNM) to analyze the data, which can effectively capture the nuanced non‐linear dynamics and lag effects in the relationship between HFMD and extreme meteorological factors. Comparing the 10th and 90th percentiles of meteorological variables with their respective median values, our results showed that extremely low temperatures and high humidity were significantly associated with increased HFMD infection risk. The greatest effect of extremely low temperatures was observed at a lag of 1–2 days, elevating the risk by 18 ∼ 33% (RR = 1.18 ∼ 1.33). Extremely high humidity was found to increase the risk of infection, starting at a lag of 4 days. In contrast, extremely high temperatures, low humidity, and high wind speed were associated with reduced risk of infection at lag of 0–12 days, with the range of RR values being 0.60–0.98 for extremely high temperatures, 0.69–0.89 for extremely low humidity, and 0.84–0.98 for extremely high wind speed respectively. Our findings suggest that extreme meteorological factors can significantly impact the incidence of HFMD in Jiangsu Province, and highlight the need for effective public health protection measures during the periods of extreme meteorological condition, particularly for vulnerable populations. Plain Language Summary Meteorological factors including temperature, humidity, rainfall and wind speed, have been recognized in previous studies as significant contributors to the spread, prevalence and severity of Hand, foot, and mouth disease (HFMD) outbreaks. In Jiangsu Province, the likelihood of experiencing such outbreaks due is high due to its dense population and substantial population movements, and it is crucial to comprehend the environmental factors driving HFMD in this region. To address this need, we collected daily HFMD infection and meteorological data from 2010 to 2017 from 13 cities in Jiangsu Province, and utilized a Distributed Lag Non‐linear Model (DLNM) to evaluate the influence of extreme meteorological factors on HFMD infections. Our findings revealed that extremely low temperatures and high humidity increased the risk of HFMD infection, while extremely high temperatures, low humidity, and high wind speed decreased the risk of HFMD infection. The results will contribute to enhancing public health preparedness and response strategies, thereby reducing the societal burden of HFMD in Jiangsu Province and safeguarding the overall health and well‐being of the population. Key Points Investigated the relationship between extreme meteorological variables and the risk of Hand, Foot, and Mouth Disease infection Quantified the effect of extreme meteorological variables using the distributed lag non‐linear model Revealed the effects of extreme meteorological variables on Hand, Foot, and Mouth Disease infections across 13 cities in Jiangsu Province

We consider the problem of testing for a dose-related effect based on a candidate set of (typically nonlinear) doseresponse models using likelihood-ratio tests. For the considered models this reduces to assessing whether the slope parameter in these nonlinear regression models is zero or not. A technical problem is that the null distribution (when the slope is zero) depends on non-identifiable parameters, so that standard asymptotic results on the distribution of the likelihood-ratio test no longer apply. Asymptotic solutions for this problem have been extensively discussed in the literature. The resulting approximations however are not of simple form and require simulation to calculate the asymptotic distribution. In addition, their appropriateness might be doubtful for the case of a small sample size. Direct simulation to approximate the null distribution is numerically unstable due to the non identifiability of some parameters. In this article, we derive a numerical algorithm to approximate the exact distribution of the likelihood-ratio test under multiple models for normally distributed data. The algorithm uses methods from differential geometry and can be used to evaluate the distribution under the null hypothesis, but also allows for power and sample size calculations. We compare the proposed testing approach to the MCP-Mod methodology and alternative methods for testing for a dose-related trend in a dose-finding example data set and simulations.