Multiple Mixed Models Research Articles

Constraining seasonal and spatial ambient and urban groundwater contributions to streamflow across a mixed land-use regional-scale Precambrian Shield watershed

Groundwater-surface water (GW-SW) interactions are complex phenomena that vary naturally over space and time and are being influenced by environmental change. The Whitson River watershed is a mixed land-use Precambrian shield watershed located in Northeastern Ontario, Canada in which groundwater is a source of municipal water supply and in which groundwater and surface waters are at risk of potential impacts from urban runoff, ongoing municipal drainage projects, periodic flooding, and climate change. This study used watershed-scale synoptic surveys of stable isotopes (δ18O and δ2H) and geochemistry, and corresponding mixing-model approaches (two-component and three-component) to quantify the seasonal and spatial variation in surface water, ambient, and urban groundwater contributions to streamflow. Multiple mixing models were generated based on isotope and geochemical source water identification, and the performance of these models was compared. Mixing-model analyses showed that groundwater contributes a critical, sustaining source to streamflow (> 50 % or more) during summer baseflow conditions, dropping to ∼ 15 % in fall when connection to shield lakes and wetlands dominate. Three-component mixing models showed similar results to the two-component models, refining groundwater contributions into ambient and urban groundwater sources. Estimated urban groundwater contributed 30 % or more to summer baseflow, dropping to ∼ 4 % in the fall. Baseflow estimates derived from graphical hydrograph separation were similar to mixing model groundwater estimates in the summer but suggest this approach overestimates groundwater contributions during the fall. These results demonstrate the complexity of source water contributions to streamflow and the challenges in their determination in mixed-land-use Precambrian Shield landscapes where communities are largely located. Assessment of source water contributions to streamflow across the Whitson River sub-watershed has generated a region-specific conceptualization (e.g. urban versus ambient groundwater) of a basin that is experiencing increased urban development, providing information that will be of value for long term management.

Using machine learning to combine genetic and environmental data for maize grain yield predictions across multi-environment trials

Key messageIncorporating feature-engineered environmental data into machine learning-based genomic prediction models is an efficient approach to indirectly model genotype-by-environment interactions.Complementing phenotypic traits and molecular markers with high-dimensional data such as climate and soil information is becoming a common practice in breeding programs. This study explored new ways to combine non-genetic information in genomic prediction models using machine learning. Using the multi-environment trial data from the Genomes To Fields initiative, different models to predict maize grain yield were adjusted using various inputs: genetic, environmental, or a combination of both, either in an additive (genetic-and-environmental; G+E) or a multiplicative (genotype-by-environment interaction; GEI) manner. When including environmental data, the mean prediction accuracy of machine learning genomic prediction models increased up to 7% over the well-established Factor Analytic Multiplicative Mixed Model among the three cross-validation scenarios evaluated. Moreover, using the G+E model was more advantageous than the GEI model given the superior, or at least comparable, prediction accuracy, the lower usage of computational memory and time, and the flexibility of accounting for interactions by construction. Our results illustrate the flexibility provided by the ML framework, particularly with feature engineering. We show that the feature engineering stage offers a viable option for envirotyping and generates valuable information for machine learning-based genomic prediction models. Furthermore, we verified that the genotype-by-environment interactions may be considered using tree-based approaches without explicitly including interactions in the model. These findings support the growing interest in merging high-dimensional genotypic and environmental data into predictive modeling.

Equivalence of variance components between standard and recursive genetic models using LDL′ transformations

BackgroundRecursive models are a category of structural equation models that propose a causal relationship between traits. These models are more parameterized than multiple trait models, and they require imposing restrictions on the parameter space to ensure statistical identification. Nevertheless, in certain situations, the likelihood of recursive models and multiple trait models are equivalent. Consequently, the estimates of variance components derived from the multiple trait mixed model can be converted into estimates under several recursive models through LDL′ or block-LDL′ transformations.ResultsThe procedure was employed on a dataset comprising five traits (birth weight—BW, weight at 90 days—W90, weight at 210 days—W210, cold carcass weight—CCW and conformation—CON) from the Pirenaica beef cattle breed. These phenotypic records were unequally distributed among 149,029 individuals and had a high percentage of missing data. The pedigree used consisted of 343,753 individuals. A Bayesian approach involving a multiple-trait mixed model was applied using a Gibbs sampler. The variance components obtained at each iteration of the Gibbs sampler were subsequently used to estimate the variance components within three distinct recursive models.ConclusionsThe LDL′ or block-LDL′ transformations applied to the variance component estimates achieved from a multiple trait mixed model enabled inference across multiple sets of recursive models, with the sole prerequisite of being likelihood equivalent. Furthermore, the aforementioned transformations simplify the handling of missing data when conducting inference within the realm of recursive models.

Association between unhealthy lifestyles and hypertension, diabetes and dyslipidemia in old adults in China

Objective: To analyze the individual and cumulative effects of unhealthy lifestyle on the prevalence of hypertension, diabetes and dyslipidemia in old adults in China, and find out the critical lifestyle in the network. Methods: Based on the baseline data of Yunnan Behavior and Disease Surveillance Cohort in 2021, a total of 16 763 older adults aged ≥60 years were included in our study. The unhealthy lifestyle factors including smoking, drinking, unhealthy eating habit, lower physical activity level, abnormal BMI and abnormal waist circumference. We calculated the unhealthy lifestyle score by using the cumulative exposures of each participant. Multiple logistic regression and mixed graphical models were used to describe the association between unhealthy lifestyle and the prevalence of hypertension, diabetes and dyslipidemia. Results: The prevalence of hypertension, diabetes and dyslipidemia were 57.0%, 11.5% and 37.0%, respectively. Most of the unhealthy lifestyles included in the study were risk factors for hypertension, diabetes and dyslipidemia, and the risks of disease increased with the increase of the unhealthy lifestyle score. The participants with the highest score (score: 6) had significantly higher prevalence of hypertension (OR=3.99, 95%CI: 1.81-8.80), diabetes (OR=4.64, 95%CI: 1.64-13.15) and dyslipidemia (OR=4.26, 95%CI: 2.08-8.73) compared with those with lowest score (score: 0). In the network constructed by mixed graphical model, abnormal waist circumference (bridge strength=0.81) and hypertension (bridge strength=0.55) were vital bridge nodes connecting unhealthy lifestyle and hypertension, diabetes and dyslipidemia. Conclusions: The unhealthy lifestyle score was associated with risks for hypertension, diabetes and dyslipidemia. Abnormal waist circumference was the key factor for chronic diseases in old adults.

Dynamic feedback algorithm based on spatial corner fitness for solving the three-dimensional multiple bin-size bin packing problem

To improve cargo loading efficiency and achieve diverse needs of companies for the loading process, this paper innovatively establishes a multiple objective mixed integer programming model for the three-dimensional multiple bin-size bin packing problem (3D-MBSBPP). The model is designed to maximize container space utilization rate and cargo load balance, minimize container usage costs, and incorporates some practical constraints. On this basis, we propose a novel dynamic feedback algorithm based on spatial corner fitness (SCF_DFA) to solve this model, which consists of three stages. Specifically, Stage 1 employs a heuristic algorithm based on spatial corner fitness to optimize the search of the remaining spaces. Stage 2 employs a container type selection algorithm to dynamically adjust and optimize container types. Stage 3 uses an improved genetic algorithm to improve the quality of the solutions of the first two stages. We demonstrate the effectiveness of the proposed algorithm through comparative experiments on benchmark instances, and apply it to solve the real-life instances for the 3D-MBSBPP. The results show that the proposed algorithm can make the average container space utilization rate reach 85.38%, which is 1.48% higher than that of baseline method, while the loading results obtained are more balanced, indicating the advantages of the SCF_DFA in solving 3D-MBSBPP. Furthermore, we conduct ablation experiments to confirm the effectiveness of each component within the algorithm.

P1130 Predictors of Remission within One Year of Advanced Therapy in Paediatric Patients with Ulcerative Colitis – analysis from Czech national registry of biologic treatment (CREdIT)

Abstract Background Data on the outcomes of advanced therapy in paediatric patients with ulcerative colitis (UC) and their predictors are scarce. We aim to assess the association between remission at one year of treatment and pre-specified predictors, including types of advanced therapy. Methods Using the Czech national prospective registry of biologic and innovative therapy of IBD (CREdIT), we identified 207 courses of treatment among paediatric UC patients. Disease remission was defined as Paediatric Ulcerative Colitis Index below 10 points. In this preliminary analysis, we employed mixed-effects regression models to attempt to identify an association between pre-specified predictors (duration of disease, first-line treatment, baseline age, baseline PUCAI, extension, ever severe, indication, baseline immunomodulator, drug) and remission at one year of the treatment. A patient, biologic center, and a time period were added as random effects into the models. Results Among 132 observations (60 female, 46%) in 108 patients, 85 were with infliximab (64%), 22 with adalimumab (17%), 20 with vedolizumab (15%), 4 with ustekinumab (3%) and 1 with tofacitinib. Most of the observations were from the first line of the biologic treatment (first: 95, 72%, second: 27, 21%, third: 7, 5.3%) and with concomitant immunosuppressive therapy at the beginning of the treatment course (97, 74%). Irrespective of treatment line, remission at 1 year was achieved in 53% (45/85) of infliximab, 36% (8/22) of adalimumab, 60% of vedolizumab and 50% of ustekinumab (2/4) treatment courses. However, we did not find a difference between the individual preparations in either the unadjusted mixed model or multiple mixed regression model. Of the pre-specified predictors, only longer duration of disease to the start of the therapy course was weakly associated with a lower remission rate in the treatment year when adjusted for treatment line (OR 1.16, 95%CI 1.00-1.34). In a sub-analysis of first-line advanced therapy only, 53% (44/83) of infliximab courses and 55% (6/11) of adalimumab courses achieved clinical remission at 1 year. Using a regression mixed model, we did not find any difference between treatment outcomes with these agents. Conclusion Based on data from real paediatric clinical practice, only half of patients with UC achieve clinical remission within a year of initiation of advanced therapy. This shows how important is to expand the drug armamentarium also in the paediatric population. Our study shows that time to initiation of advanced therapy may be an important factor even in patients with UC. Support The Grant Agency of Charles University in Prague (227023) and the Ministry of Health, Czech Republic, for the conceptual development of the research organisations (00064203).

212 Genome-Wide Association Study Investigating the Genomic Components of Efficiency in Beef Cows

Abstract Efficiency of mature beef cows consuming high-forage diets is not comparable with standard definitions of beef cattle efficiency such as feed conversion or residual feed intake in the feedlot setting. This study aimed to conduct a genome-wide association study (GWAS) to identify regions of the bovine genome associated with efficiency in mature beef cows. Ninety-eight black Angus cows were managed under extensive feeding programs over a two-year period. Using rump fat at calving, calving date, and calf weaning weight as a percentage of the body weight of the dam, a weighted percentile scoring system was used to rank efficiency. Eighty-three of the cows were retained for the GWAS. The 20 most and 20 least efficient cows (HD dataset, n=40) were genotyped with the Illumina BovineHD BeadChip (777,000 SNPs) while the remaining cows were genotyped with the Neogen GGP Bovine 100K chip (LD dataset, n = 43). The LD dataset was imputed to the HD SNP array density using Beagle5.4 (FULL dataset, n = 83). Three separate GWAS were conducted using GAPIT (version3). The first GWAS was performed on the HD dataset using a quantitative phenotype determined by the ranking system. The second GWAS was performed with the HD dataset using a qualitative phenotype (efficient vs non-efficient cows). For the final GWAS, the FULL dataset with ranked phenotypes was used. Five models were evaluated for each GWAS, including the general linear model (GLM), mixed linear model (MLM), multiple loci mixed model (MLMM), fixed and random model circulating probability unification (FarmCPU), and Bayesian-information and linkage-disequilibrium iteratively nested keyway (BLINK). The model that best represented each dataset as determined by quantile-quantile plots was used for downstream analysis. Although results of the GWAS were not significant (Bonferroni threshold), associations were identified on BTA10, 16, 17, 25, 27 and 29. Positional candidate genes in the associated regions include: FOXN3, UPB1, SNX29, DDX54, LOC112444612, RITA1, and TACC1. Further functional analyses are required to confirm these findings; however, this work provides a foundation for identifying genes and gene mutations influencing this newly described phenotype.

Effects of short-term air pollution exposure on symptoms development in the course of idiopathic pulmonary fibrosis

ABSTRACT Background Lately a potential detrimental effect of air pollution to idiopathic pulmonary fibrosis emerged. We aimed to assess the effects of short-term air pollution exposure to the clinical course of IPF. Research design and methods IPF patients were followed intensively for four nonconsecutive study periods between 13 July 2020 and 5 September 2021. Short-term exposure to O3, NO2 and PM10 concentrations was estimated using spatio-temporal land use regression models. Associations among symptoms, lung function, oxygen saturation, and short-term personal air pollutant exposure were assessed through multiple mixed effects logistic regression models. Results Data for up to 24 IPF patients (mean age: 72.2 ± 7.6 years) were analyzed. We detected positive significant associations between cough and a 10 μg/m3 increase in same day mean level of NO2 (OR = 1.59, 95%CI: 1.00–2.53), PM10 (OR = 2.42, 95%CI: 1.54–3.79), and O3 (OR = 1.63, 95%CI: 1.14–2.32). A 10 μg/m3 increase in same day mean level of NO2 was also associated with the risk of appearance of wheezing (OR = 3.01, 95%CI: 1.00–9.04), while exposure to O3 was associated with common cold (OR = 6.30, 95%CI: 3.59–11.07). No significant associations were detected between short-term exposure to air pollutants and forced vital capacity or saturation of oxygen. Conclusions Short-term exposure to increased concentrations of air pollutants is an independent risk factor for IPF symptoms' aggravation.

The Sources of Sedimentary Organic Matter Traced by Carbon and Nitrogen Isotopes and Environmental Effects during the Past 60 Years in a Shallow Steppe Lake in Northern China

The organic matter of lake sediment plays an important role in paleolimnological reconstruction. Here, we report a detailed study of organic matter components (Corg%, N%, δ13C, δ15N) in a dated sediment core of Hulun Lake in northern China. Multiple mixing models based on the stoichiometric ratios and stable isotopic compositions were applied to quantify the contributions of organic matter sources in lake sediment. The results show that the organic matter in the sediments from Hulun Lake mainly comes from terrestrial organic matter: the proportion of terrestrial organic matter is more than 80%. The results of the SIAR mixing model further reveal that the proportions of terrestrial C3 plants-derived organic matter, soil organic matter, and lake plankton-derived organic matter were 76.0%, 13.9%, and 10.1%, respectively. The organic matter content of lake sediment from terrestrial sources began to increase significantly from 1980 onward, which is consistent with the growth in overgrazing in the Hulun Lake basin. The content of organic matter from endogenous lake-derived sources began to increase significantly after 2000 due to the nutrients gradually becoming concentrated in lake water, indicating that the reduction in rivers’ discharge and the downgrade of the lake water level were the immediate causes of the lake’s environmental deterioration during this period.

Optimum: Runtime optimization for multiple mixed model deployment deep learning inference

GPUs used in data centers to perform deep learning inference tasks are underutilized. Previous systems tended to deploy a single model on a GPU to ensure that inference tasks met throughput and latency requirements. The rapid increase in one GPU’s resources, as well as the emergence of scenarios such as small models and small batches, have exacerbated the issue of low GPU utilization. In this case, a mixed model deployment-based solution can significantly improve GPU utilization while also providing greater flexibility to the inference system’s upper layer. The selection of model combinations and optimization strategies in mixed model deployment, however, remain unresolved issues. This paper proposes Optimum, the first model-combination planning and runtime optimization framework for mixed model deployment. Facing enormous search spaces, Optimum uses performance prediction for model combination selection with low search overhead. The predictor is based on a multilayer perceptron. Its input features are the profiling results of the model engine, and the output is the performance degradation. The runtime optimization strategies allow Optimum to perform performance optimization and fine-grained tradeoff. The Optimum prototype is based on CUDA multi-stream and TensorRT. The test results show that we have a flat 10.3% performance improvement over mainstream single-model deployments. We have a performance improvement of up to 7.09% over the state-of-the-art with an order of magnitude reduction in search overhead.

A quantitative approach to testing effects of different expertise levels, playing positions, and game situations on shared tactical knowledge in football teams

According to the social cognitive view, the organization of sports team behavior is underpinned by the existence of shared knowledge. This study considers tactical position knowledge as an example of shared macro-level plans involved in organizing team behavior. A coach- and a team-referenced measure of sharedness in tactical knowledge are introduced to test whether expert sport teams share knowledge to a higher degree than weaker sport teams (RQ1), whether the time spent performing a common task predicts shared knowledge (RQ2), whether role-specific specializations are mirrored in the degree to which athletes share tactical knowledge (RQ3) and whether tactical knowledge is shared to different degrees in defensive as opposed to offensive game situations (RQ4).
 Members of 17 football teams (N = 296) participating in championships at different levels of the Swiss National Football Federation took part in the scenario-based, cross-sectional study. Participants marked the positions of their team members according to their tactical playing strategy. Data were analyzed using multiple regression and mixed linear model procedures.
 Teams playing at the professional level had significantly more consistent ideas about their teams’ tactical positions than those at the amateur level. Within the amateur level, no consistent pattern of increasing consensus as the playing level increased was found. Sharedness in tactical knowledge increased with the duration of team membership when considering the team-referenced congruency measure. However, the effect was low. Overall, there was a pattern of defenders and midfielders agreeing with their team reference positions more than did goal keepers and forwards. Sharedness was generally higher in defensive situations than in offensive situations.
 The presented approach to measuring shared knowledge offers new insights into a yet under-researched field. It can be adopted to answering research questions in other sports and situations involving other measures of team performance assumed to be affected by shared knowledge.

Analysis of the Influence of Pre-Pregnancy BMI and Weight Gain during Pregnancy on the Weight of Healthy Children during the First 2 Years of Life: A Prospective Study.

Background: Increased pre-pregnancy maternal BMI (pBMI) and gestational weight gain (GWG) have been found to increase infants’ birthweight and result in the programming of child weight and impact its later weight gain. Aim: To assess the impact of pBMI and GWG on the weight of children from birth to 2 years of age and over the duration of breastfeeding. Methods: Single Centre observational prospective longitudinal cohort study. Data were collected from medical records, and medical history. The analysis of multiple linear and mixed models was involved. Findings: 20% of females were overweight, while 13% were obese before the pregnancy. An overall model, including gender and smoking, indicated a significant impact of pBMI category on a child’s birth mass (p = 0.01). The GWG category affected a child’s birth weight (p = 0.018, Effect size 0.41). pBMI did not affect the breastfeeding duration. Conclusion: pBMI and GWG correlate with birth weight and weight in neonatal period, however they become insignificant in later childhood. Weight assessment methods among children aged up to two years of age require standardization. Maternal weight before the pregnancy nor the weight gain during the pregnancy do not influence the length of breastfeeding. The biggest limitation was the small sample size and the failure to account for weight gain per trimester of pregnancy. Further research on a larger population should be continued.

Smoking and other health factors in patients with head and neck cancer

BackgroundInformation on smoking and other health factors in head and neck cancer (HNC) patients throughout treatment, follow-up and survivorship is limited. This study explores patterns of multiple health factors during radiotherapy (RT) and naturalistic long-term follow-up in a convenience sample of patients with HNC. MethodsSmoking, alcohol use and depression were measured at baseline, 4 and 12 weeks post RT for a sub-group of 99 patients who participated in a randomised controlled trial and completed long-term follow-up. These factors plus healthy eating, physical activity and fatigue are also reported from the long-term follow-up component. Smoking was measured by self-report and biochemically, whilst all other variables were by self-report. Where variables were assessed at multiple time points logistic mixed effects regression models determined within-person changes over time. ResultsThere were important discrepancies between self-reported (4–7%) and biochemically verified (13–29%) rates of smoking. Rates of smoking and hazardous alcohol intake were significantly increased at follow-up compared to baseline. Depression rates were observed to be higher at end of RT compared to baseline. At long-term follow-up, fatigue was common and co-occurred with suboptimal healthy eating and hazardous alcohol use. ConclusionClinically important levels of smoking and alcohol consumption post RT in this sample suggest possible targets for intervention beyond treatment into long-term follow-up of patients.

Application of Addition and Multiplication Noise Model Parameter Estimation in INSAR Image Processing

INSAR images are inevitably contaminated by noise during the process of generation, transmission, compression, and reception. Noise not only affects the quality of the INSAR image, but also affects subsequent operations such as the design of corresponding filters, INSAR image segmentation, compression, restoration, and feature recognition. The INSAR image noise model is mainly divided into additive noise and multiplicative noise. Compared with additive noise, multiplicative noise is more complicated due to INSAR image correlation and non-Gaussian. Based on least squares algorithm system of additive and multiplicative mixed noise model, this paper proposes a method of using PCA to remove multiplicative gamma distribution noise. The pure noise coefficient is obtained by subtracting the original coefficient from the diagonal wavelet coefficient of the noisy image, and the mode of the local variance is calculated as the estimation value of the noise standard deviation. Experiments show that the proposed method can obtain more accurate estimation of noise; in particular in the case of less noise and more detailed image information, its effect is more obvious.

Estimating soil organic carbon levels in cultivated soils from satellite image using parametric and data-driven methods

ABSTRACT Soil organic carbon (SOC) is one of the key soil components for cultivated soils. SOC is regularly monitored and mapped to improve the quality, health, and productivity of the soil. However, traditional SOC-level monitoring is expensive for land managers and farmers. Estimating SOC using satellite imagery provides an easy, efficient, and cost-effective way to monitor surface SOC levels. The objective of this study was to estimate the surface SOC distribution in selected soils of Major Land Resource Areas (MLRA), 102A (Rolling Till Plain, Brookings County, SD), and 103 (Central Iowa and Minnesota Till Prairies, Lac qui Parle County, MN), using satellite imagery with different resolutions (Landsat 8 and PlanetScope). The dominant soils in the study area are Haplustolls, Calciustolls, and Endoaquolls, which are formed in silty sediments, local silty alluvium, and till. Landsat 8 and PlanetScope spectral bands were used to develop SOC prediction models. Parametric and data-driven methods were employed to predict the SOC. Multiple linear regression and Linear Spatial Mixed Model (LSMM) were used on the Landsat 8 and PlanetScope data. In addition to the parametric models, Regression Trees and Random Forest were also employed on both satellite data. The results showed that reduced LSMM provided the lowest RMSE, which are 0.401 and 0.367 for Landsat 8 and PlanetScope, respectively. Furthermore, the random forest has the highest RPD and RPIQ for Landsat 8 (RPD of 2.67 and RPIQ of 2.49) and PlanetScope (RPD of 2.85 and RPIQ of 3.7). In all cases, models obtained from PlanetScope are better than those obtained from Landsat 8.

Remanufacturing and low-carbon investment strategies in a closed-loop supply chain under multiple carbon policies

ABSTRACT Closed-loop supply chain is an important path to realize the sustainable development of supply chain, and cap-and-trade and carbon tax policies further promote closed-loop supply chain to invest in low-carbon technologies. This paper combines Economic Production Quantity with Economic Order Quantity models and designs multiple carbon policies. Abatement rate and collection rate are the key variables used to explore the mechanism of the interaction between carbon tax and cap-and-trade, inventory dynamics of closed-loop supply chain and impact of multiple carbon policies on closed-loop supply chain. Construct and analyze benchmark model, investment model under multiple carbon policies, remanufacturing model under multiple carbon policies and mixed strategies model under multiple carbon policies, and the results show that there is a non-linear complementary relationship between carbon price and carbon tax, and both remanufacturing and low-carbon investment can optimize cost and carbon emissions when carbon price is distributed at a certain interval. Furthermore, mixed strategies combining remanufacturing and low-carbon investment can optimize cost and carbon emissions simultaneously. The main contribution of this paper is to build a low-carbon circular system, which can help balance cost and carbon emissions of closed-loop supply chain.

Performing Genome-Wide Association Studies with Multiple Models Using GAPIT.

Genome-wide association study (GWAS) is based on the linkage disequilibrium (LD) between phenotypes and genetic markers covering the whole genome. Besides the genetic linkage between the genetic markers and the causal mutations, many other factors contribute to the LD, including selection and nonrandom mating formatting population structure. Many methods have been developed with accompany of corresponding software such as multiple loci mixed model (MLMM). There are software packages that implement multiple methods to reduce the learning curve. One of them is the Genomic Association and Prediction Integrated Tool (GAPIT), which implemented eight models including GLM (General Linear Model), Mixed Linear Model (MLM), Compressed MLM, MLMM, SUPER (Settlement of mixed linear models Under Progressively Exclusive Relationship), FarmCPU (Fixed and random model Circulating Probability Unification), and BLINK (Bayesian-information and Linkage-disequilibrium Iteratively Nested Keyway). Besides the availability of multiple models, GAPIT provides comprehensive functions for data quality control, data visualization, and publication-ready quality graphic outputs, such as Manhattan plots in rectangle and circle formats, quantile-quantile (QQ) plots, principal component plots, scatter plot of minor allele frequency against GWAS signals, plots of LD between associated markers and the adjacent markers. GAPIT developers and users established a community through the GAPIT forum ( https://groups.google.com/g/gapit-forum ) with over 600 members for asking questions, making comments, and sharing experiences. In this chapter, we detail the GAPIT functions, input data frame, output files, and example codes for each GWAS model. We also interpret parameters, functional algorithms, and modules of GAPIT implementation.

Optimum: Runtime Optimization for Multiple Mixed Model Deployment Deep Learning Inference

Optimum: Runtime Optimization for Multiple Mixed Model Deployment Deep Learning Inference

Variation in Bark Allocation and Rugosity Across Seven Co-occurring Southeastern US Tree Species

Bark is a complex multifunctional structure of woody plants that varies widely among species. Thick bark is a primary trait that can protect trees from heat generated in surface fires. Outer bark on species that allocate resources to thick bark also tends to be rugose, with bark being thickest at the ridges and thinnest in the furrows. Tree diameter or wood diameter is often used as a predictor for bark thickness but little attention has been made on other factors that might affect bark development and allocation. Here we test multiple mixed effect models to evaluate additional factors (height growth rate, measure height) that correlate with bark allocation and present a method to quantify bark rugosity. We focused on seven co-occurring native tree species in the Tallahatchie Experimental Forest in north Mississippi. Approximately ten saplings of Carya tomentosa, Nyssa sylvatica, Prunus serotina, Pinus echinata, Pinus taeda, Quercus marilandica, and Quercus falcata were destructively sampled for stem analyses. Outer bark thickness (OBT) ranged from 0.01 to 0.77 cm with the thickest maximum outer bark occurring on P. taeda (0.77 cm) and the thinnest maximum outer bark occurring on P. serotina (0.17 cm). Our outer bark allocation models suggest that some individuals with rapid height growth allocate less to outer bark in C. tomentosa, N. sylvatica, P. taeda, and P. serotina, but not for P. echinata or either oak species. All species except for C. tomentosa and N. sylvatica showed evidence for outer bark taper, allocating more outer bark at the base of the bole. Inner bark also was tapered in Carya and the oaks. Bark rugosity varied among species from 0.00 (very smooth) to 0.17 (very rugose) with P. Serotina and C. tomentosa having the smoothest bark. OBT was the best fixed effect for all species. Aside from providing data for several important yet understudied species, our rugosity measures offer promise for incorporating into fluid dynamics fire behavior models.

Seroepidemiology of Crimean-Congo Hemorrhagic Fever Virus (CCHFV) in Cattle across Three Livestock Pastoral Regions in Kenya

Crimean-Congo hemorrhagic fever (CCHF) is a tick-borne zoonotic disease, endemic in Africa, with a high case fatality rate. There is no efficient treatment or licensed vaccine. This study aimed to determine the prevalence of CCHFV in cattle in extensive grazing systems (both pastoralism and ranching) within the Maasai Mara ecosystem, Nanyuki, and the Ol Pejeta Conservancy in Kenya. We conducted a seroepidemiological study of the sera of 148 cattle from 18 households from the three ecosystems in 2014, 2016, and 2019. Sera from 23 sheep and 17 goats were also obtained from the same households during the same period. Sera were analyzed for the presence of antibodies to CCHFV using the commercially available double-antigen ELISA kit. Overall, 31.5% CCHFV seropositivity was observed. The prevalence of CCHF was analyzed using a multiple logistic mixed model with main predictors. Risk factors associated with exposure to CCHFV were age (p = 0.000) and season (p = 0.007). Our findings suggest exposure to CCHFV and point to cattle as likely reservoirs of CCHFV in Kenya. The findings might play a role in providing better insights into disease risk and dynamics where analysis of tick populations in these regions should be further investigated.