Elastic Net Model Research Articles

Comparison of Linear and Non-linear Models for Coconut Yield Prediction in Coimbatore Using Weather Parameters and External Factors

Coconut is the world's most significant plantation crop, and it is grown in practically every country. As Coimbatore is the leading producer of coconut in Tamil Nadu, followed by Thanjavur and Kanyakumari, this study is centred on the Coimbatore area. West Coast Tall is a popular cultivar that produces more than other types. The West Coast Tall (WCT) cultivar was used in this research. In this paper, four models were developed such as Ridge, Least Absolute Shrinkage Selection Operator (LASSO), Elastic net (ELNET) regression methods and Artificial Neural Networks (ANN). Further, we validate this model using field-level data from TNAU coconut research farm for two years. The purpose of this communication is to find the best fit model for prediction of coconut yield using weather parameters and external factors in Coimbatore district. The models were selected based on different performance metrics such as RMSE, MAPE, MAE, and R2. Among the four models developed in the study, the ELNET model is found to be best model for prediction of coconut yield based on weather and external factors for the available data in the studied region.

Visual Field Endpoints for Neuroprotective Trials: A Case for AI-Driven Patient Enrichment

To evaluate whether an artificial intelligence (AI) model can better select candidates that would demonstrate visual field (VF) progression, in order to shorten the duration or the number of patients needed for a clinical trial. Retrospective cohort study. 7428 eyes of 3871 patients from the University of Washington Department of Ophthalmology VF Dataset were included. Progression was defined as at least 5 locations with >7 dB of change compared with baseline on 2 consecutive tests. Progression for all patients, a subgroup of the fastest progressing based on survival curves, and patients selected based on an elastic net Cox regression model were compared. The model was trained on pointwise threshold deviation values of the first VF, age, gender, laterality, and the mean total deviation (MD) at baseline. A total of 13% of all patients met the criteria for progression at 5 years. Differences in survival were observed when stratified by MD and age (P < .0001). Those at risk of progression included patients aged 60 to 80 years with an initial MD < -5.0. This subgroup decreased the sample size required to detect progression compared with the entire cohort. The AI model-selected patients required the lowest number of patients for all effect sizes and trial lengths. For a trial length of 3 years and effect size of 30%, the number of patients required was 1656 (95% CI, 1638-1674), 903 (95% CI, 884-922), and 636 (95% CI, 625-646) for the entire cohort, the subgroup, and the model-selected patients, respectively. An AI model can identify high-risk patients to substantially reduce the number of patients needed or study duration required to meet clinical trial endpoints.

RAID: Regression Analysis–Based Inductive DNA Microarray for Precise Read-Across

Chemical structure-based read-across represents a promising method for chemical toxicity evaluation without the need for animal testing; however, a chemical structure is not necessarily related to toxicity. Therefore, in vitro studies were often used for read-across reliability refinement; however, their external validity has been hindered by the gap between in vitro and in vivo conditions. Thus, we developed a virtual DNA microarray, regression analysis–based inductive DNA microarray (RAID), which quantitatively predicts in vivo gene expression profiles based on the chemical structure and/or in vitro transcriptome data. For each gene, elastic-net models were constructed using chemical descriptors and in vitro transcriptome data to predict in vivo data from in vitro data (in vitro to in vivo extrapolation; IVIVE). In feature selection, useful genes for assessing the quantitative structure–activity relationship (QSAR) and IVIVE were identified. Predicted transcriptome data derived from the RAID system reflected the in vivo gene expression profiles of characteristic hepatotoxic substances. Moreover, gene ontology and pathway analysis indicated that nuclear receptor-mediated xenobiotic response and metabolic activation are related to these gene expressions. The identified IVIVE-related genes were associated with fatty acid, xenobiotic, and drug metabolisms, indicating that in vitro studies were effective in evaluating these key events. Furthermore, validation studies revealed that chemical substances associated with these key events could be detected as hepatotoxic biosimilar substances. These results indicated that the RAID system could represent an alternative screening test for a repeated-dose toxicity test and toxicogenomics analyses. Our technology provides a critical solution for IVIVE-based read-across by considering the mode of action and chemical structures.

Candidate Plasma Biomarkers to Detect Anthracycline-Related Cardiomyopathy in Childhood Cancer Survivors: A Case Control Study in the Dutch Childhood Cancer Survivor Study.

Background Plasma biomarkers may aid in the detection of anthracycline-related cardiomyopathy (ACMP). However, the currently available biomarkers have limited diagnostic value in long-term childhood cancer survivors. This study sought to identify diagnostic plasma biomarkers for ACMP in childhood cancer survivors. Methods and Results We measured 275 plasma proteins in 28 ACMP cases with left ventricular ejection fraction <45%, 29 anthracycline-treated controls with left ventricular ejection fraction ≥53% matched on sex, time after cancer, and anthracycline dose, and 29 patients with genetically determined dilated cardiomyopathy with left ventricular ejection fraction <45%. Multivariable linear regression was used to identify differentially expressed proteins. Elastic net model, including clinical characteristics, was used to assess discrimination of proteins diagnostic for ACMP. NT-proBNP (N-terminal pro-B-type natriuretic peptide) and the inflammatory markers CCL19 (C-C motif chemokine ligands 19) and CCL20, PSPD (pulmonary surfactant protein-D), and PTN (pleiotrophin) were significantly upregulated in ACMP compared with controls. An elastic net model selected 45 proteins, including NT-proBNP, CCL19, CCL20 and PSPD, but not PTN, that discriminated ACMP cases from controls with an area under the receiver operating characteristic curve (AUC) of 0.78. This model was not superior to a model including NT-proBNP and clinical characteristics (AUC=0.75; P=0.766). However, when excluding 8 ACMP cases with heart failure, the full model was superior to that including only NT-proBNP and clinical characteristics (AUC=0.75 versus AUC=0.50; P=0.022). The same 45 proteins also showed good discrimination between dilated cardiomyopathy and controls (AUC=0.89), underscoring their association with cardiomyopathy. Conclusions We identified 3 specific inflammatory proteins as candidate plasma biomarkers for ACMP in long-term childhood cancer survivors and demonstrated protein overlap with dilated cardiomyopathy.

Comparing the Prediction Accuracy of Ridge, Lasso and Elastic Net Regression Models with Linear Regression Using Breast Cancer Data

Regularised regression methods have been developed in the past to overcome the shortcomings of ordinarily least squares (OLS) regression of not performing well with respect to both prediction accuracy and model complexity. OLS method may fail or produce regression estimates with high variance in the presence of multi-collinearity or when the predictor variables are greater than the number of observations. This study compares the predictive performance and additional information gained of Ridge, Lasso and Elastic net regularised methods with the classical OLS method using data of breast cancer patients. The findings have shown that using all the predictor variables, the OLS method failed because of the presence of multiple collinearity, while regularised Ridge, Lasso and Elastic net methods produced results that showed the predictor variables mostly significant. Using the training data, the Elastic net and Lasso seemed to indicate more significant predictor variables than the Ridge method. The result also indicated that breast cancer patients in age groups 30-39, those that are married and in stage1 of the disease, have longer survival times, while patients that are in stage2 and stage3 have shorter survival times. The OLS regression produced results only when four of the predictor variables were excluded; even then, the regularised methods still outperformed the OLS regression in terms of prediction accuracy.

벌점화 회귀분석을 이용한 English Premier League의 승점 예측모형 개발 및 경기력 요인 탐색

Advancements in Information Communication Technology during the Fourth Industrial Revolution have led to rapid progress in soccer-related industries and research. Various countries have professional soccer leagues devoted to securing victory. These activities include studies that identify the performance factors that significantly influence the outcome of matches. Researchers often use linear regression analysis based on least-squares estimation for this process, but this method has several issues. As an alternative to this, penalized regression analysis that minimizes the penalty-pointed residual sum of squares has been developed as an alternative. However, there are few studies using the penalty point regression analysis to predict points in soccer. Therefore, in this study, theoretical contents were introduced to activate the penalty point regression analysis, and further examples of application were presented by empirically analyzing 2020-2021 EPL(EPL: The Football Association Premier League Limited) data. It was found to be a statistical predictive model. According to the results, the LASSO model and elastic net models were more suitable statistical prediction models from the perspective of the root mean square error (RMSE). Additionally, among the 19 variables used in the analysis, defensive factors such as tackles, clean sheets, and saves were identified as critical performance factors that impacted the score. Based on the results of this study, in order to enable more in-depth research in the future, the researchers hope that the EPL website will provide activity data, such as the total distance traveled, number of sprints, instantaneous speed, and distance between attackers and defenders.

Hybrid machine learning models for estimating total organic carbon from mineral constituents in core samples of shale gas fields

The analysis of total organic carbon (TOC) contents is an important activity in exploring potentially hydrocarbon-generating intervals. Petroleum source rocks have, by definition, high TOC values due to the accumulation of organic matter in anoxic and low-energy deposition environments over geological time. Such petroleum generation is called conventional because it results from the increasing temperature and pressure conditions during continuous sedimentary burial. On the other hand, unconventional generation refers to reserves formed due to another heat source, for example, associated with intrusive igneous emplacement. In the last decade, there has been a growing economic interest in exploring unconventional hydrocarbon reserves because of the abundance of its occurrences. However, finding ways to assess the total organic carbon (TOC) content at lower cost and time-consuming laboratory experiments is becoming fundamental due to the importance of TOC in the analysis of the composition of these resources. This study predicts TOC using a hybrid approach, integrating machine learning models into a Grey Wolf Optimization Algorithm to adjust their parameters. The performance of four methods was compared: Elastic Net Linear Model, Extreme Learning Machine, Multivariate Adaptive Regression Splines, and Linear Support Vector Regression. The methodology was evaluated by applying core samples of the shale gas field YuDong-Nan belonging to the Sichuan Basin. The results show that the machine learning methods assisted by the evolutionary algorithm could accurately estimate TOC and be used to further exploratory geological analysis, especially those related to the prospects of unconventional oil-gas resources.

Can footwear satisfaction be predicted from mechanical properties?

Research is often conducted to investigate footwear mechanical properties and their effects on running biomechanics, but little is known about their influence on runner satisfaction, or how well the shoe is perceived. A tool to predict runner satisfaction in a shoe from its mechanical properties would be advantageous for footwear companies. Data in this study were from a database (n = 615 subject-shoe pairings) of satisfaction ratings (gathered after participants ran on a treadmill), and mechanical testing data for 87 unique subjects across 61 unique shoes. Random forest and elastic net logistic regression models were built to test if footwear mechanical properties and subject characteristics could predict runner satisfaction in 3 ways: degree-of-satisfaction on a 7-point Likert scale, overall satisfaction on a 3-point Likert scale, and willingness-to-purchase the shoe (yes/no response). Data were divided into training and validation sets, using an 80–20 split, to build the models and test their accuracy, respectively. Model accuracies were compared against the no-information rate (i.e. proportion of data belonging to the largest class). The models were not able to predict degree-of-satisfaction or overall satisfaction from footwear mechanical properties but could predict runner’s willingness to purchase with 68–75% accuracy. Midsole Gmax at the heel and forefoot appeared in the top five of variable importance rankings across both willingness-to-purchase models, suggesting its role as a major factor in purchase decisions. The negative regression coefficient for both heel and forefoot Gmax indicated that softer midsoles increase the likelihood of a shoe purchase. Future models to predict satisfaction may improve accuracy with the addition of more subject-specific parameters, such as running goals or foot proportions.

Machine Learning of Dose-Volume Histogram Parameters Predicting Overall Survival in Patients with Cervical Cancer Treated with Definitive Radiotherapy.

Purpose To analyze the effects of dosimetric parameters and clinical characteristics on overall survival (OS) by machine learning algorithms. Methods and Materials 128 patients with cervical cancer were treated with definitive pelvic radiotherapy with or without chemotherapy followed by image-guided brachytherapy. The elastic-net models with integrating DVH parameters and baseline clinical factors, only DVH parameters and only baseline clinical factors were constructed in 5-folds cross-validations for 100 iteration bootstrapping, and then were compared using concordance index (C-index) criteria. Finally, the selected important factors were used to build multivariable Cox-pH models for OS and also shown in nomograms for clinical usage. Results The median OS occurred was 25.78 months with 25 (19.53%) deaths. The elastic-net models integrating clinical and DVH factors had the best prediction performances (C-index 0.76 in the train set and C-index 0.74 in the test set). Three important factors were selected, including baseline hemoglobin level as the protective factor, primary tumor volume (GTV_P) volume, and body V5 as the risk factors. The final multivariable Cox-pH models were constructed using these important factors and had prediction performance (C-index: 0.78, 95%CI: 0.73–0.81). Conclusions This is the first attempt to establish elastic-net models to study the contributions of DVH parameters for predicting OS in patients with cervical cancer. These results can facilitate individualized tailoring of radiation treatment in cervical cancer patients.

The Gender Wealth Gap in Europe: Application of Machine Learning to Predict Individual‐level Wealth

AbstractThis article provides comparative estimates of the gender wealth gaps for 22 European countries, employing data from the Household Finance and Consumption Survey. The data on wealth are collected at the household level, while individual‐level data are needed for the estimates of gender wealth gaps. We propose a novel approach using machine learning and model averaging methods to predict individual‐level wealth data for multi‐person households. Our results suggest that random forest performs best as the predicting tool for this exercise, outperforming elastic net and Bayesian model averaging. The estimated gender wealth gaps tend to be in favor of men, especially at the top of the wealth distribution. Men have 24 percent more wealth than women on average. We also find that a high home ownership rate is associated with a smaller country‐level gender wealth gap. Our estimates suggest that the individual‐level wealth inequality is on average 3 pp higher than the household‐level wealth inequality in multi‐member households.

An epigenome-wide study of DNA methylation profiles and lung function among American Indians in the Strong Heart Study

BackgroundEpigenetic modifications, including DNA methylation (DNAm), are often related to environmental exposures, and are increasingly recognized as key processes in the pathogenesis of chronic lung disease. American Indian communities have a high burden of lung disease compared to the national average. The objective of this study was to investigate the association of DNAm and lung function in the Strong Heart Study (SHS). We conducted a cross-sectional study of American Indian adults, 45–74 years of age who participated in the SHS. DNAm was measured using the Illumina Infinium Human MethylationEPIC platform at baseline (1989–1991). Lung function was measured via spirometry, including forced expiratory volume in 1 s (FEV1) and forced vital capacity (FVC), at visit 2 (1993–1995). Airflow limitation was defined as FEV1 < 70% predicted and FEV1/FVC < 0.7, restriction was defined as FEV1/FVC > 0.7 and FVC < 80% predicted, and normal spirometry was defined as FEV1/FVC > 0.7, FEV1 > 70% predicted, FVC > 80% predicted. We used elastic-net models to select relevant CpGs for lung function and spirometry-defined lung disease. We also conducted bioinformatic analyses to evaluate the biological plausibility of the findings.ResultsAmong 1677 participants, 21.2% had spirometry-defined airflow limitation and 13.6% had spirometry-defined restrictive pattern lung function. Elastic-net models selected 1118 Differentially Methylated Positions (DMPs) as predictors of airflow limitation and 1385 for restrictive pattern lung function. A total of 12 DMPs overlapped between airflow limitation and restrictive pattern. EGFR, MAPK1 and PRPF8 genes were the most connected nodes in the protein–protein interaction network. Many of the DMPs targeted genes with biological roles related to lung function such as protein kinases.ConclusionWe found multiple differentially methylated CpG sites associated with chronic lung disease. These signals could contribute to better understand molecular mechanisms involved in lung disease, as assessed systemically, as well as to identify patterns that could be useful for diagnostic purposes. Further experimental and longitudinal studies are needed to assess whether DNA methylation has a causal role in lung disease.

Risk prediction of 30-day mortality after stroke using machine learning: a nationwide registry-based cohort study

BackgroundsWe aimed to develop and validate machine learning (ML) models for 30-day stroke mortality for mortality risk stratification and as benchmarking models for quality improvement in stroke care.MethodsData from the UK Sentinel Stroke National Audit Program between 2013 to 2019 were used. Models were developed using XGBoost, Logistic Regression (LR), LR with elastic net with/without interaction terms using 80% randomly selected admissions from 2013 to 2018, validated on the 20% remaining admissions, and temporally validated on 2019 admissions. The models were developed with 30 variables. A reference model was developed using LR and 4 variables. Performances of all models was evaluated in terms of discrimination, calibration, reclassification, Brier scores and Decision-curves.ResultsIn total, 488,497 stroke patients with a 12.3% 30-day mortality rate were included in the analysis. In 2019 temporal validation set, XGBoost model obtained the lowest Brier score (0.069 (95% CI: 0.068–0.071)) and the highest area under the ROC curve (AUC) (0.895 (95% CI: 0.891–0.900)) which outperformed LR reference model by 0.04 AUC (p < 0.001) and LR with elastic net and interaction term model by 0.003 AUC (p < 0.001). All models were perfectly calibrated for low (< 5%) and moderate risk groups (5–15%) and ≈1% underestimation for high-risk groups (> 15%). The XGBoost model reclassified 1648 (8.1%) low-risk cases by the LR reference model as being moderate or high-risk and gained the most net benefit in decision curve analysis.ConclusionsAll models with 30 variables are potentially useful as benchmarking models in stroke-care quality improvement with ML slightly outperforming others.

Using machine learning approaches to predict timely clinic attendance and the uptake of HIV/STI testing post clinic reminder messages

Timely and regular testing for HIV and sexually transmitted infections (STI) is important for controlling HIV and STI (HIV/STI) among men who have sex with men (MSM). We established multiple machine learning models (e.g., logistic regression, lasso regression, ridge regression, elastic net regression, support vector machine, k-nearest neighbour, naïve bayes, random forest, gradient boosting machine, XGBoost, and multi-layer perceptron) to predict timely (i.e., within 30 days) clinic attendance and HIV/STI testing uptake after receiving a reminder message via short message service (SMS) or email). Our study used 3044 clinic consultations among MSM within 12 months after receiving an email or SMS reminder at the Melbourne Sexual Health Centre between April 11, 2019, and April 30, 2020. About 29.5% [899/3044] were timely clinic attendance post reminder messages, and 84.6% [761/899] had HIV/STI testing. The XGBoost model performed best in predicting timely clinic attendance [mean [SD] AUC 62.8% (3.2%); F1 score 70.8% (1.2%)]. The elastic net regression model performed best in predicting HIV/STI testing within 30 days [AUC 82.7% (6.3%); F1 score 85.3% (1.8%)]. The machine learning approach is helpful in predicting timely clinic attendance and HIV/STI re-testing. Our predictive models could be incorporated into clinic websites to inform sexual health care or follow-up service.

Associations of self-reported occupational exposures and settings to ALS: a case–control study

BackgroundEnvironmental exposures contribute to the pathogenesis of amyotrophic lateral sclerosis (ALS), a fatal and progressive neurological disease. Identification of these exposures is important for targeted screening and risk factor modification.ObjectiveTo identify occupational exposures that are associated with a higher risk of ALS using both survey and standard occupational classification (SOC) coding procedures, and to highlight how exposure surveys can complement SOC coding.MethodsALS participants and neurologically healthy controls recruited in Michigan completed a detailed exposure assessment on their four most recent and longest held occupations. Exposure scores were generated from the exposure survey, and occupations were assigned to SOC codes by experienced exposure scientists.ResultsThis study included 381 ALS and 272 control participants. ALS participants reported higher duration-adjusted occupational exposure to particulate matter (OR = 1.45, 95% CI 1.19–1.78, p < 0.001), volatile organic compounds (OR = 1.22, 95% CI 1.02–1.45, p = 0.029), metals (OR = 1.48, 95% CI 1.21–1.82, p < 0.001), and combustion and diesel exhaust pollutants (OR = 1.20, 95% CI 1.01–1.43, p = 0.041) prior to ALS diagnosis, when adjusted for sex, age, and military service compared to controls. In multivariable models, only occupational exposure to metals remained significant risk (OR = 1.56, 95% CI 1.11–2.20, p = 0.011), although in an adaptive elastic net model, particulate matter (OR = 1.203), pesticides (OR = 1.015), and metals (1.334) were all selected as risk factors. Work in SOC code “Production Occupations” was associated with a higher ALS risk. SOC codes “Building and Grounds Cleaning and Maintenance Occupations”, “Construction and Extraction Occupations”, “Installation, Maintenance, and Repair Occupations”, and “Production Occupations” were all associated with a higher exposure to metals as determined using survey data.DiscussionOccupational exposure to particulate matter, volatile organic compounds, metals, pesticides, and combustion and diesel exhaust and employment in “Production Occupations” was associated with an increased ALS risk in this cohort.

Multimodal magnetic resonance neuroimaging measures characteristic of early cART-treated pediatric HIV: A feature selection approach.

Children with perinatally acquired HIV (CPHIV) have poor cognitive outcomes despite early combination antiretroviral therapy (cART). While CPHIV‐related brain alterations can be investigated separately using proton magnetic resonance spectroscopy (1H‐MRS), structural magnetic resonance imaging (sMRI), diffusion tensor imaging (DTI), and functional MRI (fMRI), a set of multimodal MRI measures characteristic of children on cART has not been previously identified. We used the embedded feature selection of a logistic elastic‐net (EN) regularization to select neuroimaging measures that distinguish CPHIV from controls and measured their classification performance via the area under the receiver operating characteristic curve (AUC) using repeated cross validation. We also wished to establish whether combining MRI modalities improved the models. In single modality analysis, sMRI volumes performed best followed by DTI, whereas individual EN models on spectroscopic, gyrification, and cortical thickness measures showed no class discrimination capability. Adding DTI and 1H‐MRS in basal measures to sMRI volumes produced the highest classification performance validation accuracy=85%AUC=0.80. The best multimodal MRI set consisted of 22 DTI and sMRI volume features, which included reduced volumes of the bilateral globus pallidus and amygdala, as well as increased mean diffusivity (MD) and radial diffusivity (RD) in the right corticospinal tract in cART‐treated CPHIV. Consistent with previous studies of CPHIV, select subcortical volumes obtained from sMRI provide reasonable discrimination between CPHIV and controls. This may give insight into neuroimaging measures that are relevant in understanding the effects of HIV on the brain, thereby providing a starting point for evaluating their link with cognitive performance in CPHIV.

Risk Factors for IUD Malposition [A36

INTRODUCTION: We aimed to identify the risk factors for intrauterine device (IUD) malposition and to develop a model to predict who is at risk for IUD malposition. METHODS: We performed a retrospective cohort study of IUD users with ultrasounds performed between July 2014 to July 2017 within a large, urban, academic medical system. Our primary outcome was IUD malposition. The Fisher’s exact test and Wilcoxon rank sum test were used. We used logistic regression, least absolute shrinkage and selection operator (LASSO), and elastic net models to predict the likelihood of having a malpositioned IUD. Institutional review board (IRB) approval was obtained for this study. RESULTS: Of 1,759 ultrasound reports with an IUD present, 436 described a malpositioned IUD. Compared to participants with non-malpositioned IUDs, participants with malpositioned IUDs were more likely to be younger (30.7 and 31.9 years old, P=.02), have fibroids (21.6% and 15.9%, P&lt;.01), and have abnormal uterine bleeding (AUB) (19.7% and 14.9%, P=.02). There were no differences between groups in gravidity, parity, uterine surgeries, number of cesarean deliveries, IUD type, or timing of IUD placement relative to abortion or delivery. Reports describing malpositioned IUDs were more likely to have been authored by radiologists than gynecologists (85.1% and 13.5%, P&lt;.01). LASSO model, elastic net model, and logistic regression were unable to identify a meaningfully predictive model for IUD malposition CONCLUSION: Fibroids and AUB are associated with IUD malposition. No other risk factors were statistically or clinically significant. No linear combination of factors meaningfully predicted risk of IUD malposition. Radiologists were more likely to author reports describing malpositioned IUDs.

Concentrations and Yields of Total Hg and MeHg in Large Boreal Rivers Linked to Water and Wetland Coverage in the Watersheds

AbstractLarge rivers are major contributors of mercury (Hg) fluxes to the ocean, as they integrate processes of loading and loss occurring at the watershed level. Stream‐scale studies have revealed that specific landscape properties, such as wetlands or lakes, are hotspots for Hg and methylmercury (MeHg) loading, sinks and transformation, but we still do not know how they operate at the whole network scale and over large geographic gradients. In this study, we evaluate how landscape metrics are related to riverine concentrations and yields of total Hg and MeHg in 18 large boreal rivers draining watersheds ranging from 44 to 209,453 km2, distributed along a 650 km latitudinal transect in the James Bay region of Québec. Our analyses of landscape metrics using elastic net models and mixed models reinforce the role of wetlands as sources of MeHg, but further show that surface coverage of water in the watershed is the major driver of both Hg and MeHg concentrations and fluxes at the whole network scale. Our findings also demonstrate that seasonality modulates the relationship between landscape properties and Hg forms. Based on hydrologic data, we additionally estimate annual fluxes for the whole Eastern James Bay to 441 kg Hg and 14.6 kg MeHg, and average landscape yield to 1.24 g Hg km−2 y−1 and 0.041 g MeHg km−2 y−1. Our study provides tools to broadly predict riverine Hg concentrations and fluxes with only a few easily accessible landscape metrics, which were shown to be better predictors than physico‐chemical variables.

Methodological evaluation of individual cognitive prediction based on the brain white matter structural connectome.

An emerging trend is to use regression‐based machine learning approaches to predict cognitive functions at the individual level from neuroimaging data. However, individual prediction models are inherently influenced by the vast options for network construction and model selection in machine learning pipelines. In particular, the brain white matter (WM) structural connectome lacks a systematic evaluation of the effects of different options in the pipeline on predictive performance. Here, we focused on the methodological evaluation of brain structural connectome‐based predictions. For network construction, we considered two parcellation schemes for defining nodes and seven strategies for defining edges. For the regression algorithms, we used eight regression models. Four cognitive domains and brain age were targeted as predictive tasks based on two independent datasets (Beijing Aging Brain Rejuvenation Initiative [BABRI]: 633 healthy older adults; Human Connectome Projects in Aging [HCP‐A]: 560 healthy older adults). Based on the results, the WM structural connectome provided a satisfying predictive ability for individual age and cognitive functions, especially for executive function and attention. Second, different parcellation schemes induce a significant difference in predictive performance. Third, prediction results from different data sets showed that dMRI with distinct acquisition parameters may plausibly result in a preference for proper fiber reconstruction algorithms and different weighting options. Finally, deep learning and Elastic‐Net models are more accurate and robust in connectome‐based predictions. Together, significant effects of different options in WM network construction and regression algorithms on the predictive performances are identified in this study, which may provide important references and guidelines to select suitable options for future studies in this field.

Development of a novel machine learning-based weighted modeling approach to incorporate Salmonella enterica heterogeneity on a genetic scale in a dose-response modeling framework.

Estimating microbial dose-response is an important aspect of a food safety risk assessment. In recent years, there has been considerable interest to advance these models with potential incorporation of gene expression data. The aim of this study was to develop a novel machine learning model that considers the weights of expression of Salmonella genes that could be associated with illness, given exposure, in hosts. Here, an elastic net-based weighted Poisson regression method was proposed to identify Salmonella enterica genes that could be significantly associated with the illness response, irrespective of serovar. The best-fit elastic net model was obtained by 10-fold cross-validation. The best-fit elastic net model identified 33 gene expression-dose interaction terms that added to the predictability of the model. Of these, nine genes associated with Salmonella metabolism and virulence were found to be significant by the best-fit Poisson regression model (p<0.05). This method could improve or redefine dose-response relationships for illness from relative proportions of significant genes from a microbial genetic dataset, which would help in refining endpoint and risk estimations.

Soil labile organic carbon indicating seasonal dynamics of soil organic carbon in northeast peatland

Peatlands have persisted as massive carbon pool in terrestrial ecosystem. Labile organic carbon (LOC) plays a vital role in carbon cycling in high-latitude peatlands, which is an important indicator of seasonal soil organic carbon (SOC) dynamics. However, the contributions of LOC fractions are poorly understood. Here, soil samples were collected at a high-latitude peatland site located in Northeastern China, and SOC and its labile fractions, including dissolved organic carbon (DOC), light-fraction organic carbon (LFOC), easily oxidizable carbon (EOC) and microbial biomass carbon (MBC) were determined, as well as the environmental and microbial parameters. The results presented seasonal variations in SOC with 352.09 ± 4.93 g/kg in July, 357.84 ± 14.02 g/kg in September, 366.74 ± 17.41 g/kg in November, 413.58 ± 7.83 g/kg in January, and 422.17 ± 8.74 g/kg in April. It was shown that DOC and LFOC were to be important indicators in seasonal carbon dynamics from an elastic net regression model. Although DOC and LFOC accounted for only 3.88% ± 0.50% and 0.34% ± 0.21% in this peatland, they might drive the changes in SOC. The redundancy analysis (RDA) model indicated that pH, total nitrogen (TN), precipitation, soil temperature and actinomycetes PLFAs are important driving factors of SOC pools. A structural equation modeling (SEM) was used to part the direct and indirect effects of factors on soil carbon dynamics based on the covariance structures. The results showed that environmental factors influenced SOC pools through different pathways. Soil pH affected carbon pools directly, while soil temperature influenced the carbon cycling indirectly by affecting soil microbes based on the path analysis. This study suggests that DOC and LFOC had the potential to cause the changes of carbon dynamics in northeast crater peatlands.