A novel prediction approach driven by graph representation learning for heavy metal concentrations.

The potential risk of heavy metals (HMs) to public health is an issue of great concern. Early prediction is an effective means to reduce the accumulation of HMs. The current prediction methods rarely take internal correlations between environmental factors into consideration, which negatively affects the accuracy of the prediction model and the interpretability of intrinsic mechanisms. Graph representation learning (GraRL) can simultaneously learn the attribute relationships between environmental factors and graph structural information. Herein, we developed the GraRL-HM method to predict the HM concentrations in soil-rice systems. The method consists of two modules, which are PeTPG and GCN-HM. In PeTPG, a graphic structure was generated using graph representation and communitization technology to explore the correlations and transmission paths of different environmental factors. Subsequently, the GCN-HM model based on the graph convolutional neural network (GCN) was used to predict the HM concentrations. The GraRL-HM method was validated by 2295 sets of data covering 21 environmental factors. The results indicated that the PeTPG model simplified correlation paths between factor nodes from 396 to 184, reducing by 53.5 % graph scale by eliminating the invalid paths. The concise and efficient graph structure enhanced the learning efficiency and representation accuracy of downstream prediction models. The GCN-HM model was superior to the four benchmark models in predicting the HM concentration in the crop, improving R2 by 36.1 %. This study develops a novel approach to improve the prediction accuracy of pollutant accumulation and provides valuable insights into intelligent regulation and planting guidance for heavy metal pollution control.

Human error identification and risk prioritization in LPG unloading operations.

Objectives. Errors due to human activities in any operation are analyzed using human reliability analysis in which the principal step is to identify potential human errors followed by quantification and analysis of the error. This work intends to apply a methodology for identifying human errors and to prioritize the risk associated with them in a liquefied petroleum gas (LPG) unloading operation. Methods. The methodology uses hierarchical task analysis which provides the basic framework, along with a systematic human error reduction and prediction approach which aids in identification and categorization of the errors associated with each task with the help of predefined error taxonomy. Also, in order to quantify the risk associated with each identified error, fuzzy failure mode and effect analysis has been adopted. To rank and prioritize the risk associated with each identified error where the individual constituent components are non-commensurable in nature, the VIseKriterijumska Optimizacija I Kompromisno Resenje method has been incorporated. Results and conclusions. Applicability of the methodology presented will help comprehend the severity of risk corresponding to each error at different levels, and the ranking mechanism thus developed in this work aids to prioritize the action to minimize the likelihood of errors.

Across-country genomic predictions in Norwegian and New Zealand Composite sheep populations with similar development history.

The goal of this study was to assess the feasibility of across-country genomic predictions in Norwegian White Sheep (NWS) and New Zealand Composite (NZC) sheep populations with similar development history. Different training populations were evaluated (i.e., including only NWS or NZC, or combining both populations). Predictions were performed using the actual phenotypes (normalized) and the single-step GBLUP via Bayesian inference. Genotyped NWS animals born in 2016 (N = 267) were used to assess the accuracy and bias of genomic estimated breeding values (GEBVs) predicted for birth weight (BW), weaning weight (WW), carcass weight (CW), EUROP carcass classification (EUC), and EUROP fat grading (EUF). The accuracy and bias of GEBVs differed across traits and training population used. For instance, the GEBV accuracies ranged from 0.13 (BW) to 0.44 (EUC) for GEBVs predicted including only NWS, from 0.06 (BW) to 0.15 (CW) when including only NZC, and from 0.10 (BW) to 0.41 (EUC) when including both NWS and NZC animals in the training population. The regression coefficients used to assess the spread of GEBVs (bias) ranged from 0.26 (BW) to 0.64 (EUF) for only NWS, 0.10 (EUC) to 0.52 (CW) for only NZC, and from 0.42 (WW) to 2.23 (EUC) for both NWS and NZC in the training population. Our findings suggest that across-country genomic predictions based on ssGBLUP might be possible for NWS and NZC, especially for novel traits.

Evolution of Reproductive Life History in Mammals and the Associated Change of Functional Constraints.

Phylogenetic trees based on multiple genomic loci enable us to estimate the evolution of functional constraints that operate on genes based on lineage-specific fluctuation of the evolutionary rate at particular gene loci, "gene-branch interactions". Using this information as predictors, our previous work inferred that the common ancestor of placental mammals was nocturnal, insectivorous, solitary, and bred seasonally. Here, we added seven new continuous traits including lifespan, bodyweight, and five reproduction-related traits and inferred the coevolution network of 14 core life history traits for 89 mammals. In this network, bodyweight and lifespan are not directly connected to each other; instead, their correlation is due to both of them coevolving with gestation period. Diurnal mammals are more likely to be monogamous than nocturnal mammals, while arboreal mammals tend to have a smaller litter size than terrestrial mammals. Coevolution between diet and the seasonal breeding behavior test shows that year-round breeding preceded the dietary change to omnivory, while seasonal breeding preceded the dietary change to carnivory. We also discuss the evolution of reproductive strategy of mammals. Genes selected as predictors were identified as well; for example, genes function as tumor suppressor were selected as predictors of weaning age.

Prospective risk assessments of patient safety events related to drug shortages in hospitals: Three actor-level perspectives.

Background: The increasing number of drug shortages (DSs) worldwide calls for more proactive solutions to prevent the negative impacts of DSs on patient care. Such solutions require in-depth knowledge about potential patient safety risks related to DSs, the processes of recognizing and managing DSs, the contextual setting in which DSs occur, and the actors involved. Objective: The aim of the study is to use prospective risk assessment to identify patient safety risks in hospitals associated with the management of DSs among actors at national, regional and local level in Denmark. Methods: Healthcare Failure Mode and Effect Analysis (HFMEA) was employed in composition with elements from the Systematic Human Error Reduction and Prediction Approach (SHERPA) and the Systems-Theoretic Accident Model and Processes (STAMP). Potential risks related to DS management across three actor levels (national, regional and local) in the Danish healthcare system were described. Each actor level consisted of six participants that were identified using a purposive sampling strategy. Processes and sub-processes related to managing critical DSs were outlined and the actors identified, prioritized and rated potential failure modes, causes and consequences related to the processes. Recommendations to mitigate failures were proposed for high risk failures modes. Results: Overall, a total of 167 failure modes were identified across the three actor levels. At the national level, the process of DS management consisted of 17 sub-processes, from which 71 failure modes were identified. Nine of them were rated as high risk. At regional level, 7 sub-processes and 33 failure modes were identified, of which 9 were rated as high risk. At local level, 14 sub-processes and 63 failure modes were identified, of which 32 were rated as high risk. The high-risk failures were related to a lack of IT support in the medication modules, underestimation of patient safety aspects, and insufficient personnel training and patient information. Conclusion: Exploring DS management failure modes across actor levels provided an overview of interrelated failures. Potential solutions related to high risk failures were developed to ensure that actors ensure patient safety related to DS in healthcare.

Identifying and reducing risks in functional endoscopic sinus surgery through a hierarchical task analysis.

OBJECTIVE: To develop a hierarchical task listing of steps required to perform successful Functional Endoscopic Sinus Surgery (FESS). To complete a technical and human factor analysis of tasks resulting in the identification of errors, frequency of occurrence, severity, and reduction through remediation. METHODS: A triangulation of methods was used in order to derive the steps required to complete a FESS: 1) a literature review was carried out of published descriptions of FESS techniques; 2) observations of three FESS; 3) interviews with surgeons on FESS techniques. Data sets were combined to develop a task analysis of a correct approach to conducting FESS. A review by 12 surgeons, and observation of 25 FESS resulted in refinement of the task analysis. With input from five consultant surgeons and one consultant anesthetist, a Systematic Human Error Reduction and Prediction Approach (SHERPA) was used to identify the risks and mitigating steps in FESS. RESULTS: Ten tasks and 49 subtasks required for a correct approach to completing FESS were identified based on literature review and expert consensus. A risk score for each subtask was calculated from a suitable risk matrix. Risk reduction methods at each subtask were detailed. High-scoring subtasks were evaluated and varying strategies examined to reduce the likelihood and mitigate the impact of error. The study demonstrates the usefulness of the HTA and SHERPA approach in standardization and optimization of clinical practice in order to improve patient safety. CONCLUSION: Hierarchical Task Analysis and SHERPA are valuable tools to deconstruct expert performance and to highlight potential errors in FESS. The HTA and SHERPA approach to surgical procedures are useful learning and assessment tools for novice surgeons. The information derived offers the opportunity to improve surgical training and enhance patient safety by identifying high-risk steps in the procedure, and how risk can be mitigated. LEVEL OF EVIDENCE: 2c Outcomes Research.

HAMDA: Hybrid Approach for MiRNA-Disease Association prediction.

For decades, enormous experimental researches have collectively indicated that microRNA (miRNA) could play indispensable roles in many critical biological processes and thus also the pathogenesis of human complex diseases. Whereas the resource and time cost required in traditional biology experiments are expensive, more and more attentions have been paid to the development of effective and feasible computational methods for predicting potential associations between disease and miRNA. In this study, we developed a computational model of Hybrid Approach for MiRNA-Disease Association prediction (HAMDA), which involved the hybrid graph-based recommendation algorithm, to reveal novel miRNA-disease associations by integrating experimentally verified miRNA-disease associations, disease semantic similarity, miRNA functional similarity, and Gaussian interaction profile kernel similarity into a recommendation algorithm. HAMDA took not only network structure and information propagation but also node attribution into consideration, resulting in a satisfactory prediction performance. Specifically, HAMDA obtained AUCs of 0.9035 and 0.8395 in the frameworks of global and local leave-one-out cross validation, respectively. Meanwhile, HAMDA also achieved good performance with AUC of 0.8965 ±â€¯0.0012 in 5-fold cross validation. Additionally, we conducted case studies about three important human cancers for performance evaluation of HAMDA. As a result, 90% (Lymphoma), 86% (Prostate Cancer) and 92% (Kidney Cancer) of top 50 predicted miRNAs were confirmed by recent experiment literature, which showed the reliable prediction ability of HAMDA.

Novel approach for predicting the joint effects based on the enzyme-catalyzed kinetics.

Organisms are exposed to mixtures of multiple contaminants and it is necessary to build prediction models for the joint effects, considering the high expense and the complexity of the traditional toxicity testing and the flood occurrence of environmental chemical pollutants. In this study, a new method for predicting the joint effects was developed and corresponding prediction models were constructed based on the kinetic models of enzyme-catalyzed reactions. While, we utilized Vibrio fischeri, Escherichia coli and Bacillus subtilis as model organisms and determined the chronic toxicity of the binary mixtures of sulfonamides (SAs) and sulfonamide potentiators (SAPs) (SA+SAP), the mixtures of two kinds of sulfonamides (SA+SA) and the binary mixtures of sulfonamide potentiators (SAPs) and tetracyclines (TCs) (SAP+TC) respectively. Finally, corresponding mixture toxicity data was utilized to fit and verify the prediction models for different joint effects.