This study aims to address accuracy challenges in assessing air pollution health impacts using Environmental Benefits Mapping and Analysis Program (BenMap), caused by limited meteorological factor data and missing pollutant data. By employing data increment strategies and multiple machine learning models, this research explores the effects of data volume, time steps, and meteorological factors on model prediction performance using several years of data from Tianjin City as an example. The findings indicate that increasing training data volume enhances the performance of Random Forest Regressor (RF) and Decision Tree Regressor (DT) models, especially for predicting CO, NO2, and PM2.5. The optimal prediction time step varies by pollutant, with the DT model achieving the highest R2 value (0.99) for CO and O3. Combining multiple meteorological factors, such as atmospheric pressure, relative humidity, and dew point temperature, significantly improves model accuracy. When using three meteorological factors, the model achieves an R2 of 0.99 for predicting CO, NO2, PM10, PM2.5, and SO2. Health impact assessments using BenMap demonstrated that the predicted all-cause mortality and specific disease mortalities were highly consistent with actual values, confirming the model\'s accuracy in assessing health impacts from air pollution. For instance, the predicted and actual all-cause mortality for PM2.5 were both 3120; for cardiovascular disease, both were 1560; and for respiratory disease, both were 780. To validate its generalizability, this method was applied to Chengdu, China, using several years of data for training and prediction of PM2.5, CO, NO2, O3, PM10, and SO2, incorporating atmospheric pressure, relative humidity, and dew point temperature. The model maintained excellent performance, confirming its broad applicability. Overall, we conclude that the machine learning and BenMap-based methods show high accuracy and reliability in predicting air pollutant concentrations and health impacts, providing a valuable reference for air pollution assessment.

Coffee Breeding programs have traditionally relied on observing plant characteristics over years, a slow and costly process. Genomic selection (GS) offers a DNA-based alternative for faster selection of superior cultivars. Stacking Ensemble Learning (SEL) combines multiple models for potentially even more accurate selection. This study explores SEL potential in coffee breeding, aiming to improve prediction accuracy for important traits [yield (YL), total number of the fruits (NF), leaf miner infestation (LM), and cercosporiosis incidence (Cer)] in Coffea Arabica. We analyzed data from 195 individuals genotyped for 21,211 single-nucleotide polymorphism (SNP) markers. To comprehensively assess model performance, we employed a cross-validation (CV) scheme. Genomic Best Linear Unbiased Prediction (GBLUP), multivariate adaptive regression splines (MARS), Quantile Random Forest (QRF), and Random Forest (RF) served as base learners. For the meta-learner within the SEL framework, various options were explored, including Ridge Regression, RF, GBLUP, and Single Average. The SEL method was able to predict the predictive ability (PA) of important traits in Coffea Arabica. SEL presented higher PA compared with those obtained for all base learner methods. The gains in PA in relation to GBLUP were 87.44% (the ratio between the PA obtained from best Stacking model and the GBLUP), 37.83%, 199.82%, and 14.59% for YL, NF, LM and Cer, respectively. Overall, SEL presents a promising approach for GS. By combining predictions from multiple models, SEL can potentially enhance the PA of GS for complex traits.

Predicting soil heavy metal (SHM) content is crucial for understanding SHM pollution levels in urban residential areas and guide efforts to reduce pollution. However, current research indicates low SHM prediction accuracy in urban areas. Therefore, we employed a deep learning method (fully connected deep neural network) alongside four other methods (muti-layer perceptron, radial basis function neural network, multiple stepwise linear regression, and Kriging interpolation) to predict SHM content in the urban residential areas of Beijing and demonstrated the strength of deep learning in improving prediction accuracy. We found the contents of the evaluated heavy metals (Cd, Cu, Pb, and Zn) exhibited significant correlations with numerous other soil physicochemical properties and environmental factors. The prediction accuracy for Cu, Pb, and Zn contents was relatively high across different methods. Notably, deep learning showed considerable strength in predicting the contents of the four heavy metals, with the R2 for the test set of the model ranging from 0.75 to 0.91. Compared to other methods, deep learning achieved markedly higher prediction accuracy according to different accuracy evaluation indicators (e.g., deep learning showed increases in the cumulative R2 of the four heavy metals ranging from 53.16 % to 187.36 % compared to other methods). Our study indicates that deep learning can significantly improve SHM content prediction accuracy in urban areas and is highly applicable in urban residential areas with complex environmental influences.

This study aims to compare the accuracy of genomic estimated breeding values (GEBV) estimated using a genomic best linear unbiased prediction (GBLUP) method and GEBV estimates incorporating prior marker information from a genome-wide association study (GWAS) for the weaning weight trait in highland Merino sheep. The objective is to provide theoretical and technical support for improving the accuracy of genomic selection. The study used a population of 1007 highland Merino ewes, with the weaning weight at 3 months as the target trait. The population was randomly divided into two groups. The first group was used for GWAS analysis to identify significant markers, and the top 5%, top 10%, top 15%, and top 20% markers were selected as prior marker information. The second group was used to estimate genetic parameters and compare the accuracy of GEBV predictions using different prior marker information. The accuracy was obtained using a five-fold cross-validation. Finally, both groups were subjected to cross-validation. The study\'s findings revealed that the heritability of the weaning weight trait, as calculated using the GBLUP model, ranged from 0.122 to 0.394, with corresponding prediction accuracies falling between 0.075 and 0.228. By incorporating prior marker information from GWAS, the heritability was enhanced to a range of 0.125 to 0.407. The inclusion of the top 5% to top 20% significant SNPs from GWAS results as prior information into GS showed potential for improving the accuracy of predicting genomic breeding value.

Genomic prediction was conducted using 2494 Japanese Black cattle from Hiroshima Prefecture and both single-nucleotide polymorphism information and phenotype data on monounsaturated fatty acid (MUFA) and oleic acid (C18:1) analyzed with gas chromatography. We compared the prediction accuracy for four models (A, additive genetic effects; AD, as for A with dominance genetic effects; ADR, as for AD with the runs of homozygosity (ROH) effects calculated by ROH-based relationship matrix; and ADF, as for AD with the ROH-based inbreeding coefficient of the linear regression). Bayesian methods were used to estimate variance components. The narrow-sense heritability estimates for MUFA and C18:1 were 0.52-0.53 and 0.57, respectively; the corresponding proportions of dominance genetic variance were 0.04-0.07 and 0.04-0.05, and the proportion of ROH variance was 0.02. The deviance information criterion values showed slight differences among the models, and the models provided similar prediction accuracy.

The genomic evaluation process relies on the assumption of linkage disequilibrium between dense single-nucleotide polymorphism (SNP) markers at the genome level and quantitative trait loci (QTL). The present study was conducted with the aim of evaluating four frequentist methods including Ridge Regression, Least Absolute Shrinkage and Selection Operator (LASSO), Elastic Net, and Genomic Best Linear Unbiased Prediction (GBLUP) and five Bayesian methods including Bayes Ridge Regression (BRR), Bayes A, Bayesian LASSO, Bayes C, and Bayes B, in genomic selection using simulation data. The difference between prediction accuracy was assessed in pairs based on statistical significance (p-value) (i.e., t test and Mann-Whitney U test) and practical significance (Cohen\'s d effect size) For this purpose, the data were simulated based on two scenarios in different marker densities (4000 and 8000, in the whole genome). The simulated data included a genome with four chromosomes, 1 Morgan each, on which 100 randomly distributed QTL and two different densities of evenly distributed SNPs (1000 and 2000), at the heritability level of 0.4, was considered. For the frequentist methods except for GBLUP, the regularization parameter λ was calculated using a five-fold cross-validation approach. For both scenarios, among the frequentist methods, the highest prediction accuracy was observed by Ridge Regression and GBLUP. The lowest and the highest bias were shown by Ridge Regression and GBLUP, respectively. Also, among the Bayesian methods, Bayes B and BRR showed the highest and lowest prediction accuracy, respectively. The lowest bias in both scenarios was registered by Bayesian LASSO and the highest bias in the first and the second scenario were shown by BRR and Bayes B, respectively. Across all the studied methods in both scenarios, the highest and the lowest accuracy were shown by Bayes B and LASSO and Elastic Net, respectively. As expected, the greatest similarity in performance was observed between GBLUP and BRR ( d = 0.007 , in the first scenario and d = 0.003 , in the second scenario). The results obtained from parametric t and non-parametric Mann-Whitney U tests were similar. In the first and second scenario, out of 36 t test between the performance of the studied methods in each scenario, 14 ( P < . 001 ) and 2 ( P < . 05 ) comparisons were significant, respectively, which indicates that with the increase in the number of predictors, the difference in the performance of different methods decreases. This was proven based on the Cohen\'s d effect size, so that with the increase in the complexity of the model, the effect size was not seen as very large. The regularization parameters in frequentist methods should be optimized by cross-validation approach before using these methods in genomic evaluation.

An appropriate flowering period is an important selection criterion in maize breeding. It plays a crucial role in the ecological adaptability of maize varieties. To explore the genetic basis of flowering time, GWAS and GS analyses were conducted using an associating panel consisting of 379 multi-parent DH lines. The DH population was phenotyped for days to tasseling (DTT), days to pollen-shedding (DTP), and days to silking (DTS) in different environments. The heritability was 82.75%, 86.09%, and 85.26% for DTT, DTP, and DTS, respectively. The GWAS analysis with the FarmCPU model identified 10 single-nucleotide polymorphisms (SNPs) distributed on chromosomes 3, 8, 9, and 10 that were significantly associated with flowering time-related traits. The GWAS analysis with the BLINK model identified seven SNPs distributed on chromosomes 1, 3, 8, 9, and 10 that were significantly associated with flowering time-related traits. Three SNPs 3_198946071, 9_146646966, and 9_152140631 showed a pleiotropic effect, indicating a significant genetic correlation between DTT, DTP, and DTS. A total of 24 candidate genes were detected. A relatively high prediction accuracy was achieved with 100 significantly associated SNPs detected from GWAS, and the optimal training population size was 70%. This study provides a better understanding of the genetic architecture of flowering time-related traits and provides an optimal strategy for GS.

BACKGROUND: Forest geneticists typically use provenances to account for population differences in their improvement schemes; however, the historical records of the imported materials might not be very precise or well-aligned with the genetic clusters derived from advanced molecular techniques. The main objective of this study was to assess the impact of marker-based population structure on genetic parameter estimates related to growth and wood properties and their trade-offs in Norway spruce, by either incorporating it as a fixed effect (model-A) or excluding it entirely from the analysis (model-B).
RESULTS: Our results indicate that models incorporating population structure significantly reduce estimates of additive genetic variance, resulting in substantial reduction of narrow-sense heritability. However, these models considerably improve prediction accuracies. This was particularly significant for growth and solid-wood properties, which showed to have the highest population genetic differentiation (QST) among the studied traits. Additionally, although the pattern of correlations remained similar across the models, their magnitude was slightly lower for models that included population structure as a fixed effect. This suggests that selection, consistently performed within populations, might be less affected by unfavourable genetic correlations compared to mass selection conducted without pedigree restrictions.
CONCLUSIONS: We conclude that the results of models properly accounting for population structure are more accurate and less biased compared to those neglecting this effect. This might have practical implications for breeders and forest managers where, decisions based on imprecise selections can pose a high risk to economic efficiency.

BACKGROUND: Identification and referral of at-risk patients from primary care practitioners (PCPs) to eye care professionals remain a challenge. Approximately 1.9 million Americans suffer from vision loss as a result of undiagnosed or untreated ophthalmic conditions. In ophthalmology, artificial intelligence (AI) is used to predict glaucoma progression, recognize diabetic retinopathy (DR), and classify ocular tumors; however, AI has not yet been used to triage primary care patients for ophthalmology referral.
OBJECTIVE: This study aimed to build and compare machine learning (ML) methods, applicable to electronic health records (EHRs) of PCPs, capable of triaging patients for referral to eye care specialists.
METHODS: Accessing the Optum deidentified EHR data set, 743,039 patients with 5 leading vision conditions (age-related macular degeneration [AMD], visually significant cataract, DR, glaucoma, or ocular surface disease [OSD]) were exact-matched on age and gender to 743,039 controls without eye conditions. Between 142 and 182 non-ophthalmic parameters per patient were input into 5 ML methods: generalized linear model, L1-regularized logistic regression, random forest, Extreme Gradient Boosting (XGBoost), and J48 decision tree. Model performance was compared for each pathology to select the most predictive algorithm. The area under the curve (AUC) was assessed for all algorithms for each outcome.
RESULTS: XGBoost demonstrated the best performance, showing, respectively, a prediction accuracy and an AUC of 78.6% (95% CI 78.3%-78.9%) and 0.878 for visually significant cataract, 77.4% (95% CI 76.7%-78.1%) and 0.858 for exudative AMD, 79.2% (95% CI 78.8%-79.6%) and 0.879 for nonexudative AMD, 72.2% (95% CI 69.9%-74.5%) and 0.803 for OSD requiring medication, 70.8% (95% CI 70.5%-71.1%) and 0.785 for glaucoma, 85.0% (95% CI 84.2%-85.8%) and 0.924 for type 1 nonproliferative diabetic retinopathy (NPDR), 82.2% (95% CI 80.4%-84.0%) and 0.911 for type 1 proliferative diabetic retinopathy (PDR), 81.3% (95% CI 81.0%-81.6%) and 0.891 for type 2 NPDR, and 82.1% (95% CI 81.3%-82.9%) and 0.900 for type 2 PDR.
CONCLUSIONS: The 5 ML methods deployed were able to successfully identify patients with elevated odds ratios (ORs), thus capable of patient triage, for ocular pathology ranging from 2.4 (95% CI 2.4-2.5) for glaucoma to 5.7 (95% CI 5.0-6.4) for type 1 NPDR, with an average OR of 3.9. The application of these models could enable PCPs to better identify and triage patients at risk for treatable ophthalmic pathology. Early identification of patients with unrecognized sight-threatening conditions may lead to earlier treatment and a reduced economic burden. More importantly, such triage may improve patients\' lives.

Determining \'excess mortality\' makes it possible to compare the burden of disasters between countries and over time, and thus also to evaluate the success of mitigation measures. However, the debate on coronavirus disease 2019 (Covid-19) has exposed that calculations of excess mortalities vary considerably depending on the method and its specification. Moreover, it is often unclear what exactly is meant by \'excess mortality\'. We define excess mortality as the excess over the number of deaths that would have been expected counter-factually, that is without the catastrophic event in question. Based on this definition, we use a very parsimonious calculation method, namely the linear extrapolation of death figures from previous years to determine the excess mortality during the Covid-19 pandemic. But unlike most other literature on this topic, we first evaluated and optimized the specification of our method using a larger historical data set in order to identify and minimize estimation errors and biases. The result shows that excess mortality rates in the literature are often inflated. Moreover, they would have exhibited considerable excess mortalities in the period before Covid-19, if this value had already been of public interest at that time. Three conclusions can be drawn from this study and its findings: (i) All calculation methods for current figures should first be evaluated against past figures. (ii) To avoid alarm fatigue, thresholds should be introduced which would differentiate between \'usual fluctuations\' and \'remarkable excess\'. (iii) Statistical offices could provide more realistic estimates.