Search

This study presents the development of a maize ear model and a predictive approach for kernel damage in maize threshing, integrating physical simulation and predictive analytics to understand better and forecast threshing-related damage. First, a maize ear model was developed to analyse kernel damage during threshing. Through the angle of repose experiments, the optimal number of spheres for the kernel model was established as 65. Tensile tests were conducted to evaluate the kernel-cob bond strength, revealing an average relative error in the bonding force of 8.71%. Vogel impact energy modelling was applied to the kernel threshing process to determine kernel damage. The correlation between the speed of seed grain movement and the occurrence of damage was analysed by post-processing to identify locations with frequent kernel damage in the drum. In-depth data analysis of kernel damage in the threshing drum further elucidates the inherent relationship between kernel velocity and damage extent. The study then focused on applying neural networks to predict damage rates. The comparative evaluation shows that the PSO-LSTM model has better prediction accuracy than LSTM and RNN models, with the PSO-LSTM network achieving an RMSE of 0.096, a R 2 of 99.96%, and a final damage rate of 2.41% in validation tests. Threshing experiments were conducted to verify the model, showing a 1.4% discrepancy between predicted and actual damage rates. This study proposes a kernel damage prediction model and provides new insights and directions for the structural design of threshing drums. • Applied Vogel model for threshing damage analysis. • Seeds are vulnerable to damage at the top and bottom of the drum. • PSO-LSTM outperforms LSTM, RNN in damage prediction. • Verified model accuracy with practical threshing experiments. • Identified optimal kernel number through repose angle tests. [ABSTRACT FROM AUTHOR]

The rational fertilisation of olive trees, based on adding exclusively the nutrients that are actually needed, is important from both the economic and environmental sustainability points of view. This paper employs UAV-obtained multispectral data collected from five different orchards located in Southern Spain to build a set of models for the prediction of the leaf nutrient content of olive trees using Support Vector Regression. The paper shows the convenience of addressing the problem as a Multiple Instance Regression, and compares two strategies of data aggregation and different choices of feature vectors derived from the raw multispectral data. The models provided good results for N, P and K (r 2 = 0.76, r 2 = 0.87 and r 2 = 0.91, respectively for the Hojiblanca model, and r 2 = 0.79, r 2 = 0.80 and r 2 = 0.80 for the Picual model). The rest of nutrients studied also offered good results for both the Picual and Hojiblanca models, ranging from r 2 = 0.69 for B to r 2 = 0.93 for Cu. The results indicate a robust performance of the models and a potential for improvement with the addition of more data, along with an advantage of considering individual models for each cultivar variety. Overall, these results are very promising for the estimation of the leaf nutrient content of olives trees and the detection of spatial variability in the fertilisation needs of orchards. [Display omitted] • Leaf nutrient content of olive trees is predicted from UAV multispectral data. • Multiple Instance Regression is applied and its benefits discussed. • Models provide good results for N, P and K in the Hojiblanca and Picual models. • The rest of nutrients also show acceptable results. • Results are promising for detecting spatial variability in fertilisation needs. [ABSTRACT FROM AUTHOR]

Predicting the plant cell response in complex environmental conditions is a challenge in plant biology. Here we developed a resource allocation model of cellular and molecular scale for the leaf photosynthetic cell of Arabidopsis thaliana , based on the Resource Balance Analysis (RBA) constraint-based modeling framework. The RBA model contains the metabolic network and the major macromolecular processes involved in the plant cell growth and survival and localized in cellular compartments. We simulated the model for varying environmental conditions of temperature, irradiance, partial pressure of CO 2 and O 2 , and compared RBA predictions to known resource distributions and quantitative phenotypic traits such as the relative growth rate, the C:N ratio, and finally to the empirical characteristics of CO 2 fixation given by the well-established Farquhar model. In comparison to other standard constraint-based modeling methods like Flux Balance Analysis, the RBA model makes accurate quantitative predictions without the need for empirical constraints. Altogether, we show that RBA significantly improves the autonomous prediction of plant cell phenotypes in complex environmental conditions, and provides mechanistic links between the genotype and the phenotype of the plant cell. • We developed a Resource Balance Analysis (RBA) model of the plant cell. • RBA accurately predicts the resource allocation of A. thaliana photosynthetic cell. • Sensitivity analyses highlighted the most influential model parameters. • RBA captures the macroscopic behavior of the Farquhar model for abiotic constraints. [ABSTRACT FROM AUTHOR]

We examine behavior in a two-stage group contest where intra-group contests are followed by an inter-group contest. Rewards accrue to the winning group, with winners of the intra-group contest within that group receiving a greater reward. The model generates a discouragement effect, where losers from the first stage exert less effort in the second stage than winners. In contrast to the related literature, we show that a prior win may be disadvantageous, generating lower profits for first stage winners as compared to losers. We consider exogenous asymmetry between groups arising from a biased group contest success function. Although the asymmetry occurs in the second stage, its effect plays out in the first stage, with higher intra-group conflict in the advantaged group. Experimental results support the qualitative predictions of the model. However, losers from the first stage bear a higher burden of the group contribution than the theoretical prediction. [ABSTRACT FROM AUTHOR]

The aim of this study is to develop a model for predicting the risk of prolonged mechanical ventilation (PMV) following surgical repair of acute type A aortic dissection (AAAD). We retrospectively collected clinical data from 381 patients with AAAD who underwent emergency surgery. Clinical features variables for predicting postoperative PMV were selected through univariate analysis, least absolute shrinkage and selection operator regression analysis, and multivariate logistic regression analysis. A risk prediction model was established using a nomogram. The model's accuracy and reliability were evaluated using the area under the curve of the receiver operating characteristic curve and the calibration curve. Internal validation of the model was performed using bootstrap resampling. The clinical applicability of the model was assessed using decision curve analysis and clinical impact curve. Among the 381 patients, 199 patients (52.2%) experienced postoperative PMV. The predictive model exhibited good discriminative ability (area under the curve = 0.827, 95% confidence interval: 0.786-0.868, P < 0.05). The calibration curve confirmed that the predicted outcomes of the model closely approximated the ideal curve, indicating agreement between the predicted and actual results (with an average absolute error of 0.01 based on 1000 bootstrap resampling). The decision curve analysis curve demonstrated that the model has significant clinical value. The nomogram model established in this study can be used to predict the risk of postoperative PMV in patients with AAAD. It serves as a practical tool to assist clinicians in adjusting treatment strategies promptly and implementing targeted therapeutic measures. [ABSTRACT FROM AUTHOR]

• MicroRNA-mediated gene regulation enables plant adaptation and survival under stressors like drought, salinity, heat, and heavy metals. • Addressing this biological importance a computational model has developed for identification of MiRNA and Pre-MiRNA based on RNA sequences. • The relative positioning based feature vector has been computed with ensemble modeling for classification purpose. • The proposed study overcome the existing studies by enhancing the accuracy of computational model. • The webservers for identification of MiRNA and Pre-MiRNA has been developed. This study proposed an intelligent model for predicting abiotic stress-responsive microRNAs in plants. MicroRNAs (miRNAs) are short RNA molecules regulates the stress in genes. Experimental methods are costly and time-consuming, as compare to in-silico prediction. Addressing this gap, the study seeks to develop an efficient computational model for plant stress response prediction. The two benchmark datasets for MiRNA and Pre-MiRNA dataset have been acquired in this study. Four ensemble approaches such as bagging, boosting, stacking, and blending have been employed. Classifiers such as Random Forest (RF), Extra Trees (ET), Ada Boost (ADB), Light Gradient Boosting Machine (LGBM), and Support Vector Machine (SVM). Stacking and Blending employed all stated classifiers as base learners and Logistic Regression (LR) as Meta Classifier. There have been a total of four types of testing used, including independent set, self-consistency, cross-validation with 5 and 10 folds, and jackknife. This study has utilized evaluation metrics such as accuracy score, specificity, sensitivity, Mathew's correlation coefficient (MCC), and AUC. Our proposed methodology has outperformed existing state of the art study in both datasets based on independent set testing. The SVM-based approach has exhibited accuracy score of 0.659 for the MiRNA dataset, which is better than the previous study. The ET classifier has surpassed the accuracy of Pre-MiRNA dataset as compared to the existing benchmark study, achieving an impressive score of 0.67. The proposed method can be used in future research to predict abiotic stresses in plants. [ABSTRACT FROM AUTHOR]

Metabolism governs cell performance in biomanufacturing, as it fuels growth and productivity. However, even in well-controlled culture systems, metabolism is dynamic, with shifting objectives and resources, thus limiting the predictive capability of mechanistic models for process design and optimization. Here, we present Cellular Objectives and State Modulation In bioreaCtors (COSMIC)-dFBA, a hybrid multi-scale modeling paradigm that accurately predicts cell density, antibody titer, and bioreactor metabolite concentration profiles. Using machine-learning, COSMIC-dFBA decomposes the instantaneous metabolite uptake and secretion rates in a bioreactor into weighted contributions from each cell state (growth or antibody-producing state) and integrates these with a genome-scale metabolic model. A major strength of COSMIC-dFBA is that it can be parameterized with only metabolite concentrations from spent media, although constraining the metabolic model with other omics data can further improve its capabilities. Using COSMIC-dFBA, we can predict the final cell density and antibody titer to within 10% of the measured data, and compared to a standard dFBA model, we found the framework showed a 90% and 72% improvement in cell density and antibody titer prediction, respectively. Thus, we demonstrate our hybrid modeling framework effectively captures cellular metabolism and expands the applicability of dFBA to model the dynamic conditions in a bioreactor. • Hybrid framework improves bioprocess model predictive capabilities. • ML-modulated stoichiometric model predicts accurate instantaneous fluxes. • Prediction accuracy of cell state shifts and resource distribution affects concentration profiles. [ABSTRACT FROM AUTHOR]

Removing senescent leaves is vital for boosting greenhouse tomato yield. This study suggests utilising sucrose concentrations in tomato petiole ends as an indicator for assessing leaf senescence. Sucrose concentrations in different leaf positions were analysed during fruiting, pre-harvesting, and harvesting periods. Leaf area and photosynthetically active radiation (PAR) data from tomato plants were input into tomato photosynthesis and respiration models simulating the yield for each compound leaf. The hyperspectral technique enabled rapid and non-destructive determination of sucrose concentration in tomato petioles using a portable spectroradiometer (ASD Fieldspec® 3) that measured spectral reflectance in 120 petiole samples. Raw spectral reflectance underwent various pretreatments. Principal component analysis (PCA) reduced data dimensions, and sucrose-spectral quantitative prediction models were established using support vector regression (SVR), k-nearest neighbour (KNN), and partial least squares (PLS). The best model for sucrose-spectral quantitative prediction was obtained by moving average (MA) preprocessing, PCA dimensionality reduction and PLS algorithm, resulting in determination coefficient of prediction set (Rp2) of 0.98 and residual predictive deviation (RPD) of 7.12. Sucrose quantitative prediction and photosynthetic respiration models exhibited a parallel trend, with assimilation capacity increasing and then decreasing with leaf position, indicating that functional leaves produce the most assimilates. Young leaves, being immature, produce limited assimilates. Despite well-developed senescent leaves, the occlusion by upper leaves hinders the release of photosynthetic potential, resulting in negative net assimilates. These findings provide compelling evidence that spectroscopy can quantitatively predict sucrose concentration in tomato petioles, aiding in effective senescent leaf identification. [Display omitted] • A quantitative relationship between spectral reflectance and sucrose was build. • Identifying senescent leaf by measuring sucrose content in tomato petioles. • Net leaf assimilation validate the senescent leaf identification criterion. [ABSTRACT FROM AUTHOR]

Bacillus species, such as Bacillus subtilis and Bacillus licheniformis , are important industrial bacteria. However, there is a lack of standardized and predictable genetic tools for convenient and reproducible assembly of genetic modules in Bacillus species to realize their full potential. In this study, we constructed a Ribosome Binding Site (RBS) library in B. licheniformis , which provides incremental regulation of expression levels over a 104-fold range. Additionally, we developed a model to quantify the resulting translation rates. We successfully demonstrated the robust expression of various target genes using the RBS library and showed that the model accurately predicts the translation rates of arbitrary coding genes. Importantly, we also extended the use of the RBS library and prediction model to B. subtilis, B. thuringiensis, and B. amyloliquefacie. The versatility of the RBS library and its prediction model enables quantification of biological behavior, facilitating reliable forward engineering of gene expression. • A shRBS library was constructed in Bacillus licheniformis , enabling incremental regulation of expression levels across a 104-fold range. • A predictive model was developed to accurately estimate the expression levels of shRBSs with arbitrary genes. • The shRBS library and its prediction model demonstrate reliability for application in other Bacillus species. • The shRBS library is feasible to regulate multiple genes and achieve desired expression levels, thereby maximizing metabolic flux towards the target products. [ABSTRACT FROM AUTHOR]

In this paper we review, upgrade, and synthesize existing models from evolutionary game theory, all of which aim at explaining human behaviour in the ultimatum game. Our new and improved versions of Gale et al. (1995) , Nowak et al. (2000) , and Rand et al. (2013) avoid shortcomings that the original versions have, one of which is that the results in the first and the last are driven by bias in the mutations. We also compare the predictions of these three models with the existing experimental evidence by looking at properties of the distributions of minimal acceptable offers. We find that the observed distributions do not conform to the predictions from Gale et al. (1995) , Rand et al. (2013) , or any other model in which there is no fitness benefit to rejecting. This does not rule out commitment-based explanations, such as Nowak et al. (2000). [ABSTRACT FROM AUTHOR]

• We propose an approach that quantitatively evaluates the impact of protein pretrained language model (PLM) in protein–ligand interaction (PLI) predictions, which allows us to select the optimal PLM for a given downstream task without exhaustively testing each PLM, thus avoiding the costly computational expense. • The mechanisms underlying the influence of protein PLMs on PLI tasks are explored. Our observations suggest that pre-training serves as a process of embedding prior knowledge, as evidenced by the increased distinctiveness of feature distributions among different tasks after pre-training. • Empirical results on three widely-used protein PLMs (TAPE, ESM-1b, and ProtTrans) across three PLI tasks (PDBbind, Kinase, and DUD-E) demonstrated that protein PLMs indeed enhanced PLI prediction, particularly in the PDBbind and Kinase tasks. Protein-ligand interaction (PLI) is a critical step for drug discovery. Recently, protein pretrained language models (PLMs) have showcased exceptional performance across a wide range of protein-related tasks. However, a significant heterogeneity exists between the PLM and PLI tasks, leading to a degree of uncertainty. In this study, we propose a method that quantitatively assesses the significance of protein PLMs in PLI prediction. Specifically, we analyze the performance of three widely-used protein PLMs (TAPE, ESM-1b, and ProtTrans) on three PLI tasks (PDBbind, Kinase, and DUD-E). The model with pre-training consistently achieves improved performance and decreased time cost, demonstrating that enhance both the accuracy and efficiency of PLI prediction. By quantitatively assessing the transferability, the optimal PLM for each PLI task is identified without the need for costly transfer experiments. Additionally, we examine the contributions of PLMs on the distribution of feature space, highlighting the improved discriminability after pre-training. Our findings provide insights into the mechanisms underlying PLMs in PLI prediction and pave the way for the design of more interpretable and accurate PLMs in the future. Code and data are freely available at https://github.com/brian-zZZ/PLM-PLI. [ABSTRACT FROM AUTHOR]

[Display omitted] When ice accumulates on a surface, it can adversely impact functionality and safety of a platform in infrastructure, transportation, and energy sectors. Despite several attempts to model the ice adhesion strength on ice-shedding materials, none have been able to justify variation in the ice adhesion strength measured by various laboratories on a simple bare substrate. This is primarily due to the fact that the effect of underlying substrate of an ice-shedding material has been entirely neglected. Here, we establish a comprehensive predictive model for ice adhesion using the shear force method on a multi-layered material. The model considers both shear resistance of the material and shear stress transfer to the underlying substrate. We conducted experiments to validate the model predictions on the effect of coating and substrate properties on the ice adhesion. The model reveals the importance of the underlying substrate of a coating on ice adhesion. Most importantly, the correlation between the ice adhesion and the coating thickness are entirely different for elastomeric and non-elastomeric materials. This model justifies different measured ice adhesion across various laboratories on the same material and elucidates how one could achieve both low ice adhesion and high mechanical durability. Such predictive model and understanding provides a rich platform to guide the future material innovation with minimal adhesion to the ice. [ABSTRACT FROM AUTHOR]

The sowing depth decision is a critical link in variable-depth sowing (VDS). Compared with rule-based sowing depth decisions, model-based methods are more intelligent and flexible under different growing conditions. This study develops a prediction model that predicts the germination and establishment of maize at different soil depths, enabling the rapid determination of the optimum sowing depth by comparing the predicted results. Dependent variables of the prediction model were selected from five seedling quality indexes, which were the emergence rate index (ER) and seedling uniformity index (U n). Independent variables were soil parameters, namely bulk density, soil depth, and soil moisture. Datasets for modelling were obtained from two experimental sites: one (240 data groups) for model training and the other (48 data groups) for model testing. After the classification of dependent variable data, a support vector machine (SVM), random forest (RF), and extreme learning machine (ELM) were used to process the training dataset and establish 186 models. Further assessment of the three selected models was performed using a test dataset. Finally, a model embedded in the SVM algorithm was developed, and achieved accuracies of 70.83% and 72.92% when predicting ER and U n. Maize emergence and establishment were effectively improved in the field by applying the developed model to select the optimum sowing depths. • A prediction model was developed to innovate research on sowing depth decision. • The model can predict the growth performance of maize at different soil depths. • Support vector machine was chosen as the modelling algorithm due to high robustness. [ABSTRACT FROM AUTHOR]

To compare performance of Assessment of Different NEoplasias in the adneXa (ADNEX model), Ovarian-Adnexal Reporting and Data System (O-RADS), Simple Rules Risk (SRR) assessment and the two-step strategy based on the application of Simple Rules (SR) followed by SRR and SR followed by ADNEX in the pre-operative discrimination between benign and malignant adnexal masses (AMs). We conducted a retrospective study from January-2018 to December-2021 in which consecutive patients with at AMs were recruited. Accuracy metrics included sensitivity (SE) and specificity (SP) with their 95% confidence intervals (CI) were calculated for ADNEX, O-RADS and SRR. When SR was inconclusive a "two-step strategy" was adopted applying SR + ADNEX model and SR + SRR assessment. A total of 514 women were included, 400 (77.8%) had a benign ovarian tumor and 114 (22.2%) had a malignant tumor. At a threshold malignancy risk of >10%, the SE and SP of ADNEX model, O-RADS and SRR were: 0.92 (95% CI, 0.86–0.96) and 0.88 (95% CI, 0.85–0.91); 0.93 (95% CI, 0.87–0.97) and 0.89 (95% CI, 0.96–0.92); 0.88 (95% CI, 0.80–0.93) and 0.84 (95% CI, 0.80–0.87), respectively. When we applied SR, 109 (21.2%) cases resulted inconclusive. The SE and SP of two-step strategy SR + SRR assessment and SR + ADNEX model were 0.88 (95% CI, 0.80–0.93) and 0.92 (95% CI, 0.89–0.94), SR + ADNEX model 0.90 (95% CI, 0.83–0.95) and 0.93 (95% CI, 0.90–0.96), respectively. O-RADS presented the highest SE, similar to ADNEX model and SR + ADNEX model. However, the SR + ADNEX model presented the higher performance accuracy with the higher SP and PPV. This two-step strategy, SR and ADNEX model applicated to inconclusive SR, is convenient for clinical evaluation. • O-RADS, ADNEX and the two-step strategy (SR plus ADENX) have a sensitivity exceeding 90% for discriminating adnexal masses. • SR plus ADNEX model applicated to inconclusive SR patients have the highest specificity and positive predictive value. • The association between SR and ADNEX model applicated to inconclusive SR patients presented the best performance accuracy [ABSTRACT FROM AUTHOR]

Adult granulosa cell tumors (AGCTs) are rare malignancies that accounts for approximately 1% of ovarian neoplasms. As there are currently no well-recognized models for predicting relapse-free survival (RFS), we performed a clinicopathological analysis to identify risk factors for AGCT recurrence. We investigated 130 patients with pathologically diagnosed AGCT as confirmed by the presence of the characteristic FOXL2 C402G mutation. Most patients had International Federation of Gynecology and Obstetrics stage I disease (n = 122, 95.3%). The 10-year RFS rate was 31.4% (22/70) and mean 10-year RFS was 74.4 (95% CI, 65.2–83.7) months. Ten patients experienced recurrence beyond the 10-year follow-up period. Undergoing fertility sparing surgery, an estrogen receptor-α (ERα) score (>0.25), and a Ki-67 index >15% were independent risk factors for recurrence in patients with stage I disease (bias-corrected C-index: 0.776). We constructed a nomogram with well-fitting calibration plots; the areas under the curve (AUCs) for 5-, and 10-year RFS prediction were 0.883 and 0.906 respectively. A simplified model with 3 predictive factors (ERα score, Ki-67 index, and primary surgical procedure) and 2 risk stratification subgroups (low- and high-risk) was constructed; its AUCs for 5-, and 10-year RFS prediction were 0.825 and 0.850 respectively. Kaplan-Meier survival curves showed significant differences in 10-year RFS between the low- and high-risk groups (p < 0.001). The type of primary surgical procedure, ERα score, and Ki-67 index are independent predictors of recurrence for patients with stage I AGCT. Our predictive model based on these factors showed good performance. • We evaluated recurrence risk in patients with adult granulosa cell tumors (AGCTs). • We focused on factors predicting recurrence in patients with stage I AGCT. • Independent predictors were the ERα score, Ki-67 index, and primary surgical procedure. • The histopathological characteristics were not associated with recurrence. • Our predictive model showed good reliability for predicting the risk of recurrence. [ABSTRACT FROM AUTHOR]

Preoperative differentiation of single-gland (SG) versus multigland (MG) primary hyperparathyroidism (PHPT) can assist with surgical planning, treatment prognostication, and patient counseling. The aim of this study was to identify preoperative predictors of SG-PHPT. Retrospective analysis of 408 patients with PHPT who underwent parathyroidectomy at a tertiary referral center. Comprehensive preoperative parameters, including demographic, laboratory, clinical, and imaging results were analyzed. Univariate analysis and binary logistic regression identified preoperative predictors of SG-PHPT. Receiver operator curves were used to analyze the predictive values of existing and novel preoperative predictive models. Elevated parathyroid hormone (PTH) (99.1 pg/mL in SG versus 93.0 pg/mL in MG), elevated calcium (10.8 mg/dL in SG versus 10.6 mg/dL in MG), lower phosphate levels (2.80 mg/dL in SG versus 2.95 mg/dL in MG), and positive imaging (ultrasound 75.6% in SG versus 56.5% in MG; sestamibi 70.8% in SG versus 45.5% in MG) were significantly associated with SG-PHPT. The Washington University Score (a predictive scoring system made from calcium, PTH, phosphate, ultrasound, and sestamibi) and the Washington University Index ([calcium × PTH]/phosphate) were comparable to previous scoring systems used to predict SG versus MG-PHPT. The association of lower phosphate with SG-PHPT is a novel finding. Previously identified predictors of SG-PHPT, including elevated PTH and positive imaging were confirmed. The Washington University Score and Index are comparable to previously described models and can be used to help surgeons predict if a patient may have SG versus MG-PHPT. [ABSTRACT FROM AUTHOR]

Cancer immunotherapy is a method of controlling and eliminating tumors by reactivating the body's cancer-immunity cycle and restoring its antitumor immune response. The increased availability of data, combined with advancements in high-performance computing and innovative artificial intelligence (AI) technology, has resulted in a rise in the use of AI in oncology research. State-of-the-art AI models for functional classification and prediction in immunotherapy research are increasingly used to support laboratory-based experiments. This review offers a glimpse of the current AI applications in immunotherapy, including neoantigen recognition, antibody design, and prediction of immunotherapy response. Advancing in this direction will result in more robust predictive models for developing better targets, drugs, and treatments, and these advancements will eventually make their way into the clinical setting, pushing AI forward in the field of precision oncology. [ABSTRACT FROM AUTHOR]

The quantum yield (QY) values of crops are important for assessing photosynthetic efficiency. Our hypothesis is that high-throughput prediction models of QY values can be established based on vegetation indices (VIs). The objectives of this research were to predict QY values based on VIs with high correlations and to compare the performances of classical machine learning with that of B-NAS methods. Compared with support vector regression (SVR) and other traditional prediction methods, in this study, a novel prediction model based on an optimised neural architecture search (B-NAS) was proposed, and achieved comparatively better results. In addition, an improved architecture was obtained and compared for ten different search architectures on the experimental dataset. The experimental results showed that the neural network model constructed based on the B-NAS method predicted QY values outperformed the models obtained using SVR and Adaboost methods in terms of all considered performance metrics (R2, RMSE, RPD, Bias). Meanwhile, the proposed model meets the criteria for a class A model. The final B-NAS model was tested on an independent validation set, obtaining similar performance metrics and an accuracy of 92% for post-prediction classification into three equal interval QY classes. An analysis of the neural architecture obtained by B-NAS showed that a four-layer network architecture had certain advantages. As well, the importance of the preprocessing and dropout layers is discussed. • Developed an optimised NAS model for predicting quantum yield of wheat in the field. • Examined the potential of NDVI, GNDVI, LCI and NDRE for prediction of quantum yield. • A B-NAS neural network with four layers and dropout normalisation gave best results. [ABSTRACT FROM AUTHOR]

Unplanned reoperation is an undesirable outcome with considerable risks and an increasingly assessed quality of care metric. There are no preoperative prediction models for reoperation after an index surgery in a broad surgical population in the literature. The Surgical Risk Preoperative Assessment System (SURPAS) preoperatively predicts 12 postoperative adverse events using 8 preoperative variables, but its ability to predict unplanned reoperation has not been assessed. This study's objective was to determine whether the SURPAS model could accurately predict unplanned reoperation. This was a retrospective analysis of the American College of Surgeons' National Surgical Quality Improvement Program adult database, 2012-2018. An unplanned reoperation was defined as any unintended operation within 30 d of an initial scheduled operation. The 8-variable SURPAS model and a 29-variable "full" model, incorporating all available American College of Surgeons' National Surgical Quality Improvement Program nonlaboratory preoperative variables, were developed using multiple logistic regression and compared using discrimination and calibration metrics: C-indices (C), Hosmer–Lemeshow observed-to-expected plots, and Brier scores (BSs). The internal chronological validation of the SURPAS model was conducted using "training" (2012-2017) and "test" (2018) datasets. Of 5,777,108 patients, 162,387 (2.81%) underwent an unplanned reoperation. The SURPAS model's C-index of 0.748 was 99.20% of that for the full model (C = 0.754). Hosmer–Lemeshow plots showed good calibration for both models and BSs were similar (BS = 0.0264, full; BS = 0.0265, SURPAS). Internal chronological validation results were similar for the training (C = 0.749, BS = 0.0268) and test (C = 0.748, BS = 0.0250) datasets. The SURPAS model accurately predicted unplanned reoperation and was internally validated. Unplanned reoperation can be integrated into the SURPAS tool to provide preoperative risk assessment of this outcome, which could aid patient risk education. • Reoperations increase morbidity and may be a quality issue. • Preoperative prediction of reoperation is informative to surgeons and patients. • 8-variable Surgical Risk Preoperative Assessment System model predicts reoperation as well as 29-variable National Surgical Quality Improvement Program model. • The Surgical Risk Preoperative Assessment System model shows good internal chronological validation. [ABSTRACT FROM AUTHOR]

There are massive jobs submitted in the supercomputer, and the job management system is typically deployed to schedule these jobs and allocate compute resources. FCFS (First Come First Serve) is a popular scheduling policy and the job's priority is determined based on the arrival time. However, under the FCFS strategy, if existing idle resources cannot meet the requirement of the head job in the waiting queue, they cannot be allocated to other jobs, which suffers from resource waste. To optimize the resource utilization, the backfilling method is proposed, which allocates the reserved idle compute nodes to a small-size and short-running non-head job, on the premise of not delaying the original head job. Obtaining the job's runtime in advance is necessary for backfilling and the traditional method relies on the user's estimation. Unfortunately, the estimated runtime provided by users is generally overestimated. Many studies extract features from historical job logs and adopt machine learning to predict the runtime. However, traditional features are insufficient to characterize the job. In this paper, we collect job logs from two supercomputers and present a novel runtime prediction framework called PREP. PREP explores the job's running path as a new feature, which implies plentiful information about the job's properties, such as the user, the project, the scale of data sets, and the parameters used. As there is a strong correlation between the job's runtime and its running path, we group jobs with similar paths into a cluster and train a runtime prediction model for each cluster respectively. Extensive evaluations demonstrate that introducing the new feature can achieve higher prediction accuracy (88.5% and 82.3% in two production supercomputers respectively), and our framework has a more desirable prediction performance than other popular strategies like Last-2 and IRPA. In addition, the predicted runtime is inserted into the real job trace of a slurm simulator to verify the advantages of PREP. • PREP explores the job's running path as a new feature to support runtime prediction. • A strong correlation between the job's runtime and its running path is evaluated. • Multiple machine learning algorithms are adopted and compared when predicting. • PREP framework suits different types of supercomputers and gain desirable prediction result. [ABSTRACT FROM AUTHOR]

Near-Infrared spectroscopy (NIR) returns full spectra in the region between 750 and 2500 nm. Although a full spectrum provides extremely informative data, sometimes this enormous amount of detail is redundant and does not bring any additional information. In this work, different attribute selection methods for the development of vineyard water status predictive models are presented. Spectra from grapevine leaves were collected on-the-go (from a moving vehicle) along nine dates during the 2015 season in a commercial vineyard using a NIR spectrometer (1200–2100 nm). Contemporarily, the stem water potential (Ψ stem) was also measured in the monitored vines. A manual selection, based on Variable Importance in Projection scores (VIP scores) to choose the spectrum intervals including the most important wavelengths (interval selection), the locally most important wavelengths in the spectrum (peak selection), as well as the Interval Partial Least Squares (IPLS) were tested as attribute selection methods. The results obtained for the estimation of Ψ stem using the whole spectrum (R2 P = 0.84, RMSEP = 0.167 MPa) were comparable to those yielded by the three attribute selection methods: the interval selection method (R2 P = 0.80, RMSEP = 0.186 MPa), the peak selection method (R2 P = 0.77, RMSEP = 0.201 MPa) and the IPLS (R2 P ∼ 0.62–0.79, RMSEP ∼ 0.186–0.252 MPa). The highest simplification was provided by two IPLS models with three wavelengths and bandwidths of 20 and 4 nm that yielded R2 P ∼0.78 and RMSEP∼ 0.190 MPa. These results corroborate the suitability of a highly reduced selection of NIR wavelengths for the prediction of grapevine water status, and its utility to develop simpler multispectral devices for vineyard water status estimation. [Display omitted] • Reduced NIR spectral information yielded accurate grapevine water status estimation. • Three different methods of NIR attribute selection were tested and compared. • Model of 1418–1422 nm, 1496–1500 nm & 1688–1692 nm worked similar to full spectrum. • NIR multispectral devices could be deployed for vineyard water status assessment. [ABSTRACT FROM AUTHOR]

Drain fluid amylase (DFA) levels have been used to predict clinically relevant postoperative pancreatic fistula (CR-POPF) and guide postoperative drain management. Optimal DFA cutoff thresholds vary between studies, thereby prompting investigation of an alternative assessment technique. As DFA measurements could, in theory, be distorted by variations in ascites fluid production, we hypothesized that adjusting DFA for volume corrected drain fluid amylase (vDFA) would improve CR-POPF predictive models. A single-institution retrospective cohort study of patients, who underwent pancreatoduodenectomies (PD) and distal pancreatectomies (DP) between 2013 and 2019, was performed. DFAs and vDFAs were measured on postoperative day (POD) 3. Clinicopathologic variables were compared between cohorts by univariable and multivariable analyses and Receiver operating characteristic (ROC) curves. Patients developing a CR-POPF were more likely to be male and have elevated DFA, vDFA, and body mass index (BMI). vDFA use did not contribute to a superior CR-POPF predictive model compared to DFA–a finding consistent on subanalysis of surgery type PD versus DP. In CR-POPF predictive models, DFA, vDFA, and male sex significantly improved CR-POPF predictive models when considering both surgery subtypes, while only DFA and vDFA significantly improved models when cohorts were segregated by surgery type. Postoperative DFA remains a preferred method of predicting CR-POPF as the proposed vDFA assessment technique only adds complexity without increased discriminability. [ABSTRACT FROM AUTHOR]

Postoperative pneumonia (POP) is a common complication following lung cancer surgery and is associated with increased hospitalization costs and mortalities. We aimed to identify risk factors associated with POP and to develop a reliable predictive model. Patients who underwent lung cancer surgery between January 2015 and December 2021 in our hospital were enrolled. Least absolute shrinkage and selection operator regression analysis was used to select predictors of POP. Multivariable logistic regression was performed to construct the nomogram. Bootstrap resampling was conducted for internal validation. The performance of the model was evaluated by discrimination and calibration. A total of 5269 consecutive patients were enrolled. POP occurred in 1.7% of patients (92/5269). Five independent predictors were identified: age, predicted forced expiratory volume in 1 s, predicted diffusing capacity of the lungs for carbon monoxide, tuberculosis history, and surgery duration. The multivariable regression model showed good discrimination (C-index: 0.821, 95% confidence interval, 0.783-0.859), which was well validated by internal validation. The calibration curve illustrated good agreement between the predicted probability and observed probability of POP. Based on the easily available risk factors, our nomogram could predict the risk of POP with good discrimination and calibration. The model has good clinical practicability, enabling precise and targeted interventions to reduce the incidence of POP in high-risk patients. [ABSTRACT FROM AUTHOR]

We investigate self-serving belief distortion about dominant norms of honesty. Consider an environment where the subject can earn a monetary reward by lying. In contrast to the existing literature on motivated beliefs, we do not focus on distortion in one dimension alone, but instead consider beliefs in two dimensions: empirical (what other people do) and normative (what other people approve of). Our experimental findings are consistent with the predictions of a dual-self model in which conditional norm-followers strategically distort their beliefs to justify self-serving behavior. We argue that the asymmetry between what we infer from empirical as opposed to normative information is a key ingredient of belief distortion in our context: widespread honest behavior is a strong indicator of disapproval of lying (and thus that a norm of honesty is followed), but the opposite does not hold. Taken together, we show why, when, and which norm-relevant beliefs are strategically distorted. [ABSTRACT FROM AUTHOR]

To evaluate AERMOD for predicting livestock odour dispersion in the Canadian climate and flat terrain, 22 odour plumes were measured around a broiler barn for two days with distinct atmospheric stability classes, namely Aug 31st with C & D and Sep 2nd with A & B. The diurnally varying odour emission rates were measured and input for the modelling. The relationship between odour concentration (OC) and intensity (OI) previously established for this broiler barn was utilised to convert the measured OI to OC or modelled OC to OI. For comparing the hourly modelled predictions and 10-min field measurements, two peak-to-mean ratios (PTM) including a constant factor of 4 (PTM C) and a variable factor dependent on the atmospheric stability (PTM V) were applied. However, neither of them could sufficiently diminish the significance in the paired differences between model predictions and field measurements (p < 0.05). The efficacy of scaling factors specifically developed for the corresponding atmospheric stability classes based on the PTM adjustments were demonstrated, leading to an average OI agreement of 76% for Aug 31st but 97% for Sep 2nd, regardless of the PTM type. The choice of scaling factors could significantly influence the model performance. For Aug 31st, SF Geo was favourable for improving model predictions over short distances (within 300 or 400 m), otherwise SF All should be used for guaranteeing insignificant paired differences. In contrast, SF Geo was consistently preferred for Sep 2nd in terms of improving or at least retaining the acceptable percentage of FB i besides narrowing the paired differences. • AERMOD was assessed for predicting livestock odour dispersion in a cold flat region. • Previously obtained diurnal OE and OC–OI relationship of the broiler barn was used. • Constant or atmospheric stability based PTM was poor in adjusting model accuracy. • Scaling factor should be specifically derived for each atmospheric stability class. • Average OI agreement was 76% for Aug 31st (C & D) but 97% for Sep 2nd (A & B). [ABSTRACT FROM AUTHOR]

Generally, the computational grid consists of a large number of computing nodes, some of them are idle due to the uneven geographical distribution of computing requirements. This may cause workload unbalancing problems, which affect the performance of large-scale computational grids. In order to balance the computing requirements and computing nodes, we propose a job scheduling algorithm based on the workload prediction of computing nodes. We first analyze the causes of workload imbalance and the feasibility of reallocating computing resources. Secondly, we design an application and workload-aware scheduling algorithm (AWAS) by combining the previously designed workload prediction model. To reduce the complexity of the AWAS algorithm, we propose a parallel job scheduling method based on computing node workload prediction. The experiments show that the AWAS algorithm can balance the workload among different computing nodes on the real-world dataset. In addition, we propose the parallelism of workload prediction model from the perspective of internal structure and data set to make AWAS apply to more computing nodes of the large-scale computing grids. Experimental results show that the combination of the two can achieve satisfactory acceleration efficiency. • In this paper, we propose an application and grid computing nodes workload aware scheduling algorithm (AWAS). • AWAS optimizes grid loads using greedy and heuristic methods. • This paper parallelizes the scheduling process of AWAS to improve its efficiency. • The experiment results show that the AWAS can achieve workload balancing in China National Grid. [ABSTRACT FROM AUTHOR]

Reducing energy consumption and optimizing resource usage in large cloud data centers is still an essential target for the current researchers and cloud providers. The state-of-the-art highlights the effectiveness of VM consolidation and live migrations in achieving reasonable solutions. However, most proposals consider only the real-time workload variations to decide whether a host is overloaded or underloaded, or to trigger migration actions. Such approaches may apply frequent and needless VM migrations leading to energy waste, performance degradation, and service-level agreement (SLA) violations. In this paper, we propose a consolidation approach based on the resource utilization prediction to determine the overloaded and underloaded hosts. The prediction method combines a Kalman filter and support vector regression (SVR) to forecast the host's future CPU utilization. Simulations are conducted on Cloudsim using real PlanetLab workloads to verify the performance of our proposal against existing benchmark algorithms. Experimental results demonstrate that our consolidation technique significantly reduces the SLA violation rate, number of VM migrations, and energy consumed in the datacenter. • A multi-step workload prediction model based on Kalman Filter and SVR is proposed. • A dynamic VM consolidation approach based on resource prediction is presented. • A trade-off between SLA violation and energy consumption is pursued. • Analysis study is carried out on the energy consumed by executing our simulation. • Several experiments are conducted on Cloudsim to evaluate our approach. [ABSTRACT FROM AUTHOR]

Hyperspectral imaging has widely been used for plant phenotyping applications. Calibration transfer is crucial for eliminating the prediction drift when a model developed in one plant phenotyping system needs to be deployed in another system. In this study two approaches (Domain Adversarial Neural Network - DANN and Adversarial Discriminative Domain Adaptation – ADDA) for transferring a deep learning model across hyperspectral imaging-based plant phenotyping systems were presented. The proposed techniques employ adversarial learning to minimise the discrepancy between the spectral data of two imaging facilities in the feature space. The calibration transfer analysis was examined by collecting the hyperspectral images of maize plants in two phenotyping facilities equipped with different sensors covering the visible near-infrared (VNIR) region under different lighting conditions. The capability of the suggested methods to transfer a relative water content (RWC) prediction model was investigated. The results showed that the prediction R2 improved from 0.603 to 0.849 and 0.843 and RMSE reduced from 23.52% to 14.15% and 12.73% for DANN and ADDA, respectively. The study showed that both approaches improved the performance of the transferred model compared to a similar deep learning model without adversarial transfer. Visualisation of feature space showed that both methods lead to similar source and target feature distributions, thus indicating that a prediction model developed using these common features can successfully be applied to both imaging facilities. The main benefit of the DANN and ADDA is that they do not need the standard samples to be imaged in two phenotyping systems or the collection of labelled data to carry out the transfer. The adversarial learning-based calibration transfer can promote the sharing of prediction models across different phenotyping systems. • Differences across phenotyping facilities hinders the application of exiting models. • Calibration transfer between different phenotyping facilities is critically important. • Adversarial domain adaptation based methods were proposed for calibration transfer. • Proposed methods resulted in RWC model shared between two facilities. • Proposed methods can help to eliminate the need of full recalibration. [ABSTRACT FROM AUTHOR]

In this paper, we present a highly accurate performance estimation methodology that accounts for architecture slack in workloads. Our work leverages the advanced instrumentation available in POWER8 processor that monitors core pipeline activity in relation to off-core memory accesses to build metrics for architecture slack characterization for workloads. Using these metrics, we construct a workload classifier that classifies workloads as core-bound and memory-bound and propose a performance prediction model for change in processor frequency for each class of workload – cPerf and mPerf, respectively. We evaluated these models with SPECCPU and PARSEC benchmark suites on a POWER8 based OpenPOWER system. We observed that the predicted performance with our models has high accuracy (97%) for both CPU and memory intensive benchmarks. We validated that the classifier is suitable to accurately classify phase of workloads during execution intervals. We developed an algorithm that uses classifier for phase classification and prediction models for performance estimation at runtime. We leveraged this algorithm and evaluated the execution time impacts of CPU and memory classified benchmarks. • DVFS for multi-threaded applications using core pipeline metrics. • Core pipeline metrics to accurately classify compute or memory bound workloads. • Monitoring activity of core pipeline units for architecture slack estimation. • Architecture slack for accurate phase classification and performance estimation. [ABSTRACT FROM AUTHOR]

The accurate prediction of indoor air temperatures is vital for the timely control of thermal environments in livestock buildings. However, ensuring the accuracy of air temperature predictions is challenging due to the uncertainty introduced by external factors and periodic fluctuations. New approaches are required to improve the prediction accuracy of models for indoor air temperatures. A discrete model is developed that incorporates time-period groups (TPGs), the group buffered rolling (GBR) mechanism, and TPG factors. The measured temperature data were divided into four TPGs in chronological order, and the GBR mechanism with TPG factors eliminated fluctuations and periodicity from the time-series data. The efficacy of the proposed model was verified by comparing its results with those of two existing models (the grey model and discrete grey model) and filed experiments in layer hen houses. The traditional model only reflected the increasing trend of original air temperature data, but did not reflect periodic fluctuations. The mean absolute percentage error (APE) was >10% in the traditional model. The GBR mechanism and TPG factors allowed the new model to reflect deviations in the original time-series at each time point in real-time, thus minimizing the APE and improving the prediction performance. The modified model was superior to the traditional models, exhibited higher accuracy and reliable performance (APE < 10%), and accurately captured the periodic fluctuations in air temperature. It is a helpful tool for predicting indoor air temperatures, and may help in the development of improved control strategies for the indoor thermal environment. • A new discrete grey model modified with the group buffered rolling mechanism. • The GBR-DGM shows improved prediction performance for indoor air temperature. • Accurately predicts temporal fluctuations in air temperatures in a layer hen house. • Applications include the development of improved environmental control systems. [ABSTRACT FROM AUTHOR]

Portable mid-infrared spectroscopy (pMIRS) combined with machine learning was used to predict selected parameters for soil organic carbon (SOC) storage. In particular, SOC, soil inorganic C (SIC), hot-water extractable C (hwC), clay and sand content were predicted for ten vineyards with varying geopedological settings. As a pre-test, spectra were collected from sieved and pressed tablets with 30 and 90 kPa respectively and compared to powdery samples in order to optimise sample preparation. Further, spectra from 30 kPa tablets were used to calibrate prediction models for a sample set (n = 540) of 10 vineyards with pronounced geopedological variation using Support Vector Machines (SVM). The calibrated SVM models performed well with R2 = 0.81–0.98 and RPIQ = 5.20–13.0 for all investigated parameters. Third, two years after the calibration samples, follow-up samples were collected from four of the vineyards. While the models performed excellent for hwC (R2 = 0.93), prediction accuracy for SOC was lower. Segmentation of the total dataset into SIC-free and SIC-containing samples resulted in better predictions of SOC of the first sampling period. For the prediction of the follow-up sampling dates, model performance could not be maintained. We conclude that pMIRS-SVM calibrations are suited for the prediction of parameters related to soil C storage under varying geopedological conditions and may provide potential for future C monitoring. Extending the database with additional samples from geopedological scenarios not included in this dataset may strengthen model robustness and help to evaluate effects of SIC content on model performance. • Compressed soil tablets reduce sample preparation labour for portable mid-infrared spectroscopy. • SVM model calibrations provide robust predictions for parameters related SOC storage under varying geopedological conditions. • Prediction for follow-up sampling via calibrated SVM models highlights the potential for future SOC monitoring purposes. • Arbitrarily occurring SIC influences MIR SVM model performance. [ABSTRACT FROM AUTHOR]

To report performance of an integrated predictive model (IPM) algorithm based on patient factors, surgical resectability and surgical complexity to predict outcome of primary cytoreductive surgery (PCS) and guide treatment plan in patients with advanced epithelial ovarian cancer (AEOC). Patients with AEOC between October 2018 and October 2020 were enrolled into a dedicated AEOC program and decision for PCS or neoadjuvant chemotherapy (NACT) was based on multidisciplinary consensus. Data of unresectable stage IVb, patient factors (PF), surgical resectability scores (SRS) and surgical complexity scores (SCS) was prospectively documented. An integrated prediction model (IPM) was developed to predict outcome of optimal (RD < 1 cm) cytoreduction. Retrospective analysis was performed to assess the performance of the IPM. Cut-offs were selected using the Youden Index. Of 185 eligible patients, 81 underwent PCS and 104 were treated with NACT. Patients undergoing PCS had significantly lower median PF (0 vs 2, p < 0.01), SRS (2 vs 4, p < 0.01) and pre-operative SCS (6 vs 8.5, p = 0.01) compared to NACT. In patients undergoing PCS, 88% had optimal cytoreduction and 34.5% had grade 3–4 post-operative complications. A model triaging patients with unresectable Stage IVb, PF > 2, SRS > 5 and SCS > 9 to NACT had 85% sensitivity, 75% specificity and 85% accuracy for outcome of optimal cytoreduction. Our model would have improved triage of 3/10 sub-optimally cytoreduced patients to NACT. For outcome of no-gross residual disease (RD = 0 mm) using the same cut-offs sensitivity and specificity were 85% and 76% respectively. The 4-step IPM algorithm had high sensitivity and specificity for optimal cytoreduction with acceptable morbidity without delay to adjuvant therapy. This algorithm may be used to triage patients to PCS or NACT once it is further validated. • Decision between PCS and NACT is often subjective and may correlate with survival. • We developed a 4-step IPM factoring stage IVb unresectable, Patient factors, Surgical resectability and Surgical complexity. • The model had 85% sensitivity, 75% specificity and 85% accuracy for outcome of optimal cytoreduction at PCS. • Similar performance was obtained for outcome of no gross residual. • This model would have triaged 3 suboptimally cytoreduced patients to NACT and improves precision in decision making [ABSTRACT FROM AUTHOR]

• Multiple cortical features fusion via an transformer-based classifier. • Achieved high AUC in classifying stable and progressive MCI. • The caudal anterior cingulate and pars orbitalis contributed most to the classification task. • VOL and CT are more reliable biomarkers than CD, FD, and SA in the progression of MCI. Mild cognitive impairment (MCI) is usually considered the early stage of Alzheimer's disease (AD). Therefore, the accurate identification of MCI individuals with high risk in converting to AD is essential for the potential prevention and treatment of AD. Recently, the great success of deep learning has sparked interest in applying deep learning to neuroimaging field. However, deep learning techniques are prone to overfitting since available neuroimaging datasets are not sufficiently large. Therefore, we proposed a deep learning model fusing cortical features to address the issue of fusion and classification blocks. To validate the effectiveness of the proposed model, we compared seven different models on the same dataset in the literature. The results show that our proposed model outperformed the competing models in the prediction of MCI conversion with an accuracy of 83.3% in the testing dataset. Subsequently, we used deep learning to characterize the contribution of brain regions and different cortical features to MCI progression. The results revealed that the caudal anterior cingulate and pars orbitalis contributed most to the classification task, and our model pays more attention to volume features and cortical thickness features. [ABSTRACT FROM AUTHOR]

• Biological markers did not improve machine learning prediction of clinical outcome in CHR. • Complement proteins (Factor X, C1r subcomponent, C4A & C5) associate inversely with functional outcome. • C 5 associate positively with positive symptoms severity. Functional outcomes are important measures in the overall clinical course of psychosis and individuals at clinical high-risk (CHR), however, prediction of functional outcome remains difficult based on clinical information alone. In the first part of this study, we evaluated whether a combination of biological and clinical variables could predict future functional outcome in CHR individuals. The complement and coagulation pathways have previously been identified as being of relevance to the pathophysiology of psychosis and have been found to contribute to the prediction of clinical outcome in CHR participants. Hence, in the second part we extended the analysis to evaluate specifically the relationship of complement and coagulation proteins with psychotic symptoms and functional outcome in CHR. We carried out plasma proteomics and measured plasma cytokine levels, and erythrocyte membrane fatty acid levels in a sub-sample (n = 158) from the NEURAPRO clinical trial at baseline and 6 months follow up. Functional outcome was measured using Social and Occupational Functional assessment Score (SOFAS) scale. Firstly, we used support vector machine learning techniques to develop predictive models for functional outcome at 12 months. Secondly, we developed linear regression models to understand the association between 6-month follow-up levels of complement and coagulation proteins with 6-month follow-up measures of positive symptoms summary (PSS) scores and functional outcome. A prediction model based on clinical and biological data including the plasma proteome, erythrocyte fatty acids and cytokines, poorly predicted functional outcome at 12 months follow-up in CHR participants. In linear regression models, four complement and coagulation proteins (coagulation protein X, Complement C1r subcomponent like protein, Complement C4A & Complement C5) indicated a significant association with functional outcome; and two proteins (coagulation factor IX and complement C5) positively associated with the PSS score. Our study does not provide support for the utility of cytokines, proteomic or fatty acid data for prediction of functional outcomes in individuals at high-risk for psychosis. However, the association of complement protein levels with clinical outcome suggests a role for the complement system and the activity of its related pathway in the functional impairment and positive symptom severity of CHR patients. [ABSTRACT FROM AUTHOR]

Floods result in substantial damage throughout the world every year. Accurate predictions of floods can significantly alleviate casualties and property losses. However, due to the complexity of hydrology process especially in a city with complicated pipe network, the accuracy of traditional flood forecasting models suffer from the performance degradation with the increasing of required prediction period. In the work, based on the collected historical data of Xixian City, Henan Province, China, using the Internet of Things system (IoT) in 2011-2018, a Bidirectional Gated Recurrent Unit (BiGRU) multi-step flood prediction model with attention mechanism is proposed. In our model, the attention mechanism is used to automatically adjust the matching degree between the input features and output. Besides, we use a bidirectional GRU model, which can process the input sequence from two directions of time series (chronologically and antichronologically), then merge their representations together. Compared with the prediction model using Long Short Term Memory (LSTM), our method can generate better prediction result, as can be seen from the arrival time error and peak error of floods during multi-step predictions. • A BiGRU multi-step flood prediction model based on attention mechanism was proposed. • Model tests were designed and compared with relevant models based on LSTM neural network. • The results show that the model has smaller deterministic coefficient and RMSE. • Within the allowable error range, the prediction accuracy of our model can reach 100%. [ABSTRACT FROM AUTHOR]

• The first genomic machine learning web server based on genomic coordinates. • Convenient online genomic feature extraction with respect to genomic intervals. • High-accuracy predictive model construction and automated web deployment. • Considerably high accuracy achieved in terms of Sensitivity, Specificity, and AUC. The primary sequences of DNA, RNA and protein have been used as the dominant information source of existing machine learning tools, especially for contexts not fully explored by wet-experimental approaches. Since molecular markers are profoundly orchestrated in the living organisms, those markers that cannot be unambiguously recovered from the primary sequence often help to predict other biological events. To the best of our knowledge, there is no current tool to build and deploy machine learning models that consider genomic evidence. We therefore developed the WHISTLE server, the first machine learning platform based on genomic coordinates. It features convenient covariate extraction and model web deployment with 46 distinct genomic features integrated along with the conventional sequence features. We showed that, when predicting m6A sites from SRAMP project, the model integrating genomic features substantially outperformed those based on only sequence features. The WHISTLE server should be a useful tool for studying biological attributes specifically associated with genomic coordinates, and is freely accessible at: www.xjtlu.edu.cn/biologicalsciences/whi2. [ABSTRACT FROM AUTHOR]

• A novel method was proposed to identify RNA m2G sites using hybrid features. • The over-sample method SMOTE was adopted to deal with the problem of data imbalance. • After using MRMD to select features, the performance of the model is improved. • The RFhy-m2G is superior to other methods, which can effective identify m2G sites. N2-methylguanosine is a post-transcriptional modification of RNA that is found in eukaryotes and archaea. The biological function of m2G modification discovered so far is to control and stabilize the three-dimensional structure of tRNA and the dynamic barrier of reverse transcription. To discover additional biological functions of m2G, it is necessary to develop time-saving and labor-saving calculation tools to identify m2G. In this paper, based on hybrid features and a random forest, a novel predictor, RFhy-m2G, was developed to identify the m2G modification sites for three species. The hybrid feature used by the predictor is used to fuse the three features of ENAC, PseDNC, and NPPS. These three features include primary sequence derivation properties, physicochemical properties, and position-specific properties. Since there are redundant features in hybrid features, MRMD2.0 is used for optimal feature selection. Through feature analysis, it is found that the optimal hybrid features obtained still contain three kinds of properties, and the hybrid features can more accurately identify m2G modification sites and improve prediction performance. Based on five-fold cross-validation and independent testing to evaluate the prediction model, the accuracies obtained were 0.9982 and 0.9417, respectively. The robustness of the predictor is demonstrated by comparisons with other predictors. [ABSTRACT FROM AUTHOR]

[Display omitted] • Considering both RNA and protein expression could obtain more biological evidence. • We propose a method for predicting protein expression based on the diffusion model. • The diffusion model is the first to be used in the field of single-cell analysis. • The proposed method was validated on three single-cell sequencing datasets. • Our results provide new directions for the identification of novel drug targets. The executors of organismal functions are proteins, and the transition from RNA to protein is subject to post-transcriptional regulation; therefore, considering both RNA and surface protein expression simultaneously can provide additional evidence of biological processes. Cellular indexing of transcriptomes and epitopes by sequencing (CITE-Seq) technology can measure both RNA and protein expression in single cells, but these experiments are expensive and time-consuming. Due to the lack of computational tools for predicting surface proteins, we used datasets obtained with CITE-seq technology to design a deep generative prediction method based on diffusion models and to find biological discoveries through the prediction results. In our method, the scDM, which predicts protein expression values from RNA expression values of individual cells, uses a novel way of encoding the data into a model and generates predicted samples by introducing Gaussian noise to gradually remove the noise to learn the data distribution during the modelling process. Comprehensive evaluation across different datasets demonstrated that our predictions yielded satisfactory results and further demonstrated the effectiveness of incorporating information from single-cell multiomics data into diffusion models for biological studies. We also found that new directions for discovering therapeutic drug targets could be provided by jointly analysing the predictive value of surface protein expression and cancer cell drug scores. [ABSTRACT FROM AUTHOR]

RNA modification, N4-acetylcytidine (ac4C), is enzymatically catalyzed by N-acetyltransferase 10 (NAT10) and plays an essential role across tRNA, rRNA, and mRNA. It influences various cellular functions, including mRNA stability and rRNA biosynthesis. Wet-lab detection of ac4C modification sites is highly resource-intensive and costly. Therefore, various machine learning and deep learning techniques have been employed for computational detection of ac4C modification sites. The known ac4C modification sites are limited for training an accurate and stable prediction model. This study introduces GANSamples-ac4C, a novel framework that synergizes transfer learning and generative adversarial network (GAN) to generate synthetic RNA sequences to train a better ac4C modification site prediction model. Comparative analysis reveals that GANSamples-ac4C outperforms existing state-of-the-art methods in identifying ac4C sites. Moreover, our result underscores the potential of synthetic data in mitigating the issue of data scarcity for biological sequence prediction tasks. Another major advantage of GANSamples-ac4C is its interpretable decision logic. Multi-faceted interpretability analyses detect key regions in the ac4C sequences influencing the discriminating decision between positive and negative samples, a pronounced enrichment of G in this region, and ac4C-associated motifs. These findings may offer novel insights for ac4C research. The GANSamples-ac4C framework and its source code are publicly accessible at http://www.healthinformaticslab.org/supp/. [Display omitted] • A prediction tool with interpretable predictive logic is proposed for RNA modification N4-acetylcytidine (ac4C) sites. • The proposed tool GANSamples-ac4C utilizes transfer learning and GAN. • GAN-synthesized samples can mitigate the issue of data scarcity. • GANSamples-ac4C outperforms the existing tools of ac4C prediction. • The code is released to facilitate the GAN-synthesized samples for similar tasks. [ABSTRACT FROM AUTHOR]

In this article, we present an edge-based architecture for enhancing the autonomous capabilities of resource-constrained aerial robots by enabling a remote nonlinear model predictive control scheme, which can be computationally heavy to run on the aerial robots' onboard processors. The nonlinear model predictive control is used to control the trajectory of an unmanned aerial vehicle while detecting, and preventing potential collisions. The proposed edge architecture enables trajectory recalculation for resource-constrained unmanned aerial vehicles in relatively real-time, which will allow them to have fully autonomous behaviors. The architecture is implemented with a remote Kubernetes cluster on the edge side, and it is evaluated on an unmanned aerial vehicle as our controllable robot, while the robotic operating system is used for managing the source codes, and overall communication. With the utilization of edge computing and the architecture presented in this work, we can overcome computational limitations, that resource-constrained robots have, and provide or improve features that are essential for autonomous missions. At the same time, we can minimize the relative travel time delays for time-critical missions over the edge, in comparison to the cloud. We investigate the validity of this hypothesis by evaluating the system's behavior through a series of experiments by utilizing either the unmanned aerial vehicle or the edge resources for the collision avoidance mission. • This paper presents an edge computing architecture for controlling the trajectory of edge connected robots. • The proposed approach utilizes the benefits of edge computing to enable optimal behavior and collision-free trajectories. • The complete framework is introduced and experimentally evaluated. [ABSTRACT FROM AUTHOR]

With the advent of the second industrial revolution, mining and metallurgical processes generate large volumes of tailings and mine wastes (TMW), which worsens global environmental pollution. Studying the occurrence of metal and metalloid elements in TMW is an effective approach to evaluating pollution linked to TMW. However, traditional laboratory-based measurements are complicated and time-consuming; thus, an empirical method is urgently needed that can rapidly and accurately determine elemental occurrence forms. In this study, a model combining Bayesian optimization and random forest (RF) approaches was proposed to predict TMW occurrence forms. To build the RF model, a dataset of 2376 samples was obtained, with mineral composition, elemental properties, and total concentration composition used as inputs and the percentage of occurrence forms as the model output. The correlation coefficient (R), coefficient of determination, mean absolute error, root mean squared error, and root mean squared logarithmic error metrics were used for model evaluation. After Bayesian optimization, the optimal RF model achieved accurate predictive performance, with R values of 0.99 and 0.965 on the training and test sets, respectively. The feature significance was analyzed using feature importance and Shapley additive explanatory values, which revealed that the electronegativity and total concentration of the elements were the two features with the greatest influence on the model output. As the electronegativity of an element increases, its corresponding residual fraction content gradually decreases. This is because the solubility typically increases with the solvent's polarity and electronegativity. Overall, this study proposes an RF model based on the nature of TMW that can rapidly and accurately predict the percentage values of metal and metalloid element occurrence forms in TMW. This method can minimize testing time requirements and help to assess TMW pollution risks, as well as further promote safe TMW management and recycling. [Display omitted] • A model was proposed for the elemental occurrence forms of tailings and mine wastes (TMW). • The elemental occurrence forms in TMW can be predicted quickly and accurately. • Electronegativity and total concentration were the most important variables. • The influential mechanism of input features on occurrence forms was studied. [ABSTRACT FROM AUTHOR]

The effectiveness of deep learning models is greatly dependent on the availability of a vast amount of labeled data. However, in the realm of surface defect classification, acquiring and annotating defect samples proves to be quite challenging. Consequently, accurately predicting defect types with only a limited number of labeled samples has emerged as a prominent research focus in recent years. Few-shot learning, which leverages a restricted sample set in the support set, can effectively predict the categories of unlabeled samples in the query set. This approach is particularly well-suited for defect classification scenarios. In this article, we propose a transductive few-shot surface defect classification method, which using both the instance-level relations and distribution-level relations in each few-shot learning task. Furthermore, we calculate class center features in transductive manner and incorporate them into the feature aggregation operation to rectify the positioning of edge samples in the mapping space. This adjustment aims to minimize the distance between samples of the same category, thereby mitigating the influence of unlabeled samples at category boundary on classification accuracy. Experimental results on the public dataset show the outstanding performance of our proposed approach compared to the state-of-the-art methods in the few-shot learning settings. Our code is available at https://github.com/Harry10459/CIDnet. • Data shortage leads to model performance degradation. • Improve model performance in case of sample shortage. • Construction of sample distribution-level relationships. • Solutions to the problem of model overfitting. • Class center features improve classification performance. [ABSTRACT FROM AUTHOR]

Blade geometric variations generally have a significant impact on the structural dynamics of integrally bladed disks widely used in the advanced aero-engines. This paper presents a holistic research in regard to the predictive dynamic modeling, analysis and experimental verification for a blisk by taking advantage of the advanced 3D optical geometry measurement technology. Geometrically mistuned models (GMMs) are semi-automatically constructed upon the precisely measured blisk geometry by an efficient FE mesh updating strategy. They provide explicit, high-fidelity digital representations of the geometric variations within the integrally manufactured blisk. A 'Sector Mode Assembling Reduction Technique' is developed and specifically tailored for efficient dynamic analysis of the large-sized GMMs at a relatively low computational cost and memory requirement. Intensive test campaigns, including forced response tests in the stationary/spinning rig under well-controlled laboratory conditions, are carried out for a full assessment of the GMMs' dynamic prediction capability. Experimental verification results show that the GMM is able to capture the modal dynamics and resonant vibration of the stationary/rotating blisk with satisfactory accuracy. The physical-reality-based GMM converted directly from the precise geometry measurement data can be considered as a viable and valuable tool for predictive vibration evaluation of blisks. However, its model accuracy exhibits a mode-related dependence on the mesh density. The tradeoff between model accuracy and prohibitive computational cost proved to be the bottleneck of this promising blisk modeling approach. • Geometrically mistuned models (GMMs) are constructed upon measured blisk geometry. • GMMs are dynamically reduced by a Sector Mode Assembling Reduction Technique. • GMM's accuracy exhibits a mode-related dependence on the mesh density. • GMMs are validated by forced response tests in the stationary/spinning rig. • GMM captures resonant vibration of the stationary/rotating blisk with good accuracy. [ABSTRACT FROM AUTHOR]

The United States Environmental Protection Agency (USEPA) uses the lethal dose 50% (LD 50) value from in vivo rat acute oral toxicity studies for pesticide product label precautionary statements and environmental risk assessment (RA). The Collaborative Acute Toxicity Modeling Suite (CATMoS) is a quantitative structure-activity relationship (QSAR)-based in silico approach to predict rat acute oral toxicity that has the potential to reduce animal use when registering a new pesticide technical grade active ingredient (TGAI). This analysis compared LD 50 values predicted by CATMoS to empirical values from in vivo studies for the TGAIs of 177 conventional pesticides. The accuracy and reliability of the model predictions were assessed relative to the empirical data in terms of USEPA acute oral toxicity categories and discrete LD 50 values for each chemical. CATMoS was most reliable at placing pesticide TGAIs in acute toxicity categories III (>500–5000 mg/kg) and IV (>5000 mg/kg), with 88% categorical concordance for 165 chemicals with empirical in vivo LD 50 values ≥ 500 mg/kg. When considering an LD 50 for RA, CATMoS predictions of 2000 mg/kg and higher were found to agree with empirical values from limit tests (i.e. , single, high-dose tests) or definitive results over 2000 mg/kg with few exceptions. • The CATMoS model was evaluated for predicting the rat LD 50 of new pesticide technical grade active ingredients (TGAI). • Predictions were compared to empirical rat acute toxicity data of 177 TGAIs registered by the USEPA from 1998 to 2020. • Model reliability was high when placing TGAIs in USEPA acute toxicity categories III and IV. • CATMoS discrete value predictions of 2000 mg/kg and higher were found to agree with empirical values with few exceptions. • Our evaluation supports the potential use of CATMoS predictions of TGAI acute oral toxicity in place of animal studies. [ABSTRACT FROM AUTHOR]

In Bangladesh, cholera poses a significant environmental health risk. Yet, information about the severity of cholera risk is limited as risk varies over time and changing weather patterns make historical cholera risk predictions less reliable. In this paper, we examine how households use geographically and temporally personalized cholera risk predictions to inform their beliefs and behaviors related to cholera and its aversion. We estimate how access to a smartphone application containing monthly cholera risk predictions unique to a user's home location affects households' beliefs about their cholera risk and their water use and hygiene behaviors. We find that households with access to this application feel more equipped to respond to environmental and health risks and reduce their reliance on surface water for bathing and washing – a common cholera transmission pathway. We do not find that households invest additional resources into drinking water treatment, nor do we find reductions in self-reported cholera incidence. Further, households with a static, non-personalized app containing public health information about cholera exhibit similar patterns of beliefs updating. Taken together, our results suggest that access to risk information can help households make safer water choices, yet improving design and credibility remain important dimensions for increasing application usability. • Satellite observations provide improved information for cholera risk. • Households in rural Bangladesh find improved information for cholera risk valuable. • Dynamic, personalized cholera risk information can change household water behaviors. • Households with a cholera early warning system reduce reliance on surface water. • Early warning system design affects sustained application use. [ABSTRACT FROM AUTHOR]

The rise of antimicrobial resistance (AMR) is one of the most relevant problems for human and animal health. According to One Health Approach, it is important to regulate the use of antimicrobials and monitor the spread of AMR in the environment as well. Apis mellifera (L. 1758) colonies were used as bioindicators thanks to their physical and behavioural characteristics. During their foraging flights, bees can intercept small particles, including atmospheric particulate matter, etc., and also microorganisms. To date, the antimicrobial surveillance network is limited to the sanitary level but lacks into environmental context. This study aimed to evaluate the use of A. mellifera colonies distributed throughout the Emilia-Romagna region (Italy) as indicators of environmental antimicrobial-resistant bacteria. This was performed by creating a statistical predictive model that establishes correlations between environmental characteristics and the likelihood of isolating specific bacterial genera and antimicrobial-resistant strains. A total of 608 strains were isolated and tested for susceptibility to 19 different antimicrobials. Aztreonam-resistant strains were significantly related to environments with sanitary structures, agricultural areas and wetlands, while urban areas present a higher probability of trimethoprim/sulfamethoxazole-resistant strains isolation. Concerning genera, environments with sanitary structures and wetlands are significantly related to the genera Proteus spp., while the Escherichia spp. strains can be probably isolated in industrial environments. The obtained models showed maximum values of Models Accuracy and robustness (R2) of 55 % and 24 %, respectively. The results indicate the efficacy of utilizing A. mellifera colonies as valuable bioindicators for estimating the prevalence of AMR in environmentally disseminated bacteria. This survey can be considered a good basis for the development of further studies focused on monitoring both sanitary and animal pathology, creating a specific network in the environments of interest. [Display omitted] • The antimicrobial resistance rise is a relevant problem for human and animal health. • Honey bees (Apis mellifera L.) can contribute to monitoring antimicrobial resistance. • Antimicrobial-resistant bacteria depend on agriculture fragmentation and land use. • Statistical predictive model correlates AMR honey bee bacteria to land use. • The AMR spread can be statistically predicted with agriculture and land use. [ABSTRACT FROM AUTHOR]

• Seizure recurrence prediction tools have not been tested in IGE patients. • Seizures recurred in half of our patients, 68% relapsed within the first year. • A seizure-free period of 2 years was the only factor associated with epilepsy remission. • Larger studies are needed to predict seizure recurrence in IGE patients. Currently, there is a limited availability of tools to predict seizure recurrence after discontinuation of antiseizure medications (ASMs). This study aimed to establish the seizure recurrence rate following ASM cessation in adult patients with idiopathic generalized epilepsy (IGE) and to assess the predictive performance of the Lamberink and the Stevelink prediction models using real-world data. Retrospective longitudinal study in IGE patients who underwent ASM withdrawal in a tertiary epilepsy clinic since June 2011, with the latest follow up in January 2024. The minimum follow-up period was 12 months. Clinical and demographic variables were collected, and the seizure recurrence prediction models proposed by Lamberink and Stevelink were applied and evaluated. Forty-seven patients (mean age 33.15 ± 8 [20–55] years; 72.35 % women) were included. During the follow-up period, seizures recurred in 25 patients (53.2 %). Median time to recurrence was 8 months [IQR 3–13.5 months], and 17 patients (68 %) relapsed within the first year. None of the relapsing patients developed drug-resistant epilepsy. The only significant risk factor associated with recurrence was a seizure-free period of less than 2 years before discontinuing medication (91.7 % vs 40 %, p =.005). The Stevelink prediction model at both 2 (p =.015) and 5 years (p =.020) achieved statistical significance, with an AUC of 0.72 (95 % CI 0.56–0.88), while the Lamberink model showed inadequate prognostic capability. In our real-world cohort, a seizure-free period of at least 2 years was the only factor significantly associated with epilepsy remission after ASM withdrawal. Larger studies are needed to accurately predict seizure recurrence in IGE patients. [ABSTRACT FROM AUTHOR]

[S U M M A R Y] Many miRNA-disease association prediction models incorporate Gaussian interaction profile kernel similarity (GIPS). However, the GIPS fails to consider the specificity of the miRNA-disease association matrix, where matrix elements with a value of 0 represent miRNA and disease relationships that have not been discovered yet. To address this issue and better account for the impact of known and unknown miRNA-disease associations on similarity, we propose a method called vector projection similarity-based method for miRNA-disease association prediction (VPSMDA). In VPSMDA, we introduce three projection rules and combined with logistic functions for the miRNA-disease association matrix and propose a vector projection similarity measure for miRNAs and diseases. By integrating the vector projection similarity matrix with the original one, we obtain the improved miRNA and disease similarity matrix. Additionally, we construct a weight matrix using different numbers of neighbors to reduce the noise in the similarity matrix. In performance evaluation, both LOOCV and 5-fold CV experiments demonstrate that VPSMDA outperforms seven other state-of-the-art methods in AUC. Furthermore, in a case study, VPSMDA successfully predicted 10, 9, and 10 out of the top 10 associations for three important human diseases, respectively, and these predictions were confirmed by recent biomedical resources. [Display omitted] • Three projection rules are proposed to make full use of false negative data. • The vector projection similarity is constructed by combining three projection rules and logical functions. • The weight matrices for miRNA and disease similarity were constructed with varying neighborhood sizes. [ABSTRACT FROM AUTHOR]

• Distractor suppression (DS) was measured using visual search task. • Using functional connectivity between large anatomical regions to predict DS. • Prefrontal-temporal, insula-limbic, parietal-occipital connections were key to DS. • The model could also predict daily distractibility in an independent sample. Distractor suppression (DS) is crucial in goal-oriented behaviors, referring to the ability to suppress irrelevant information. Current evidence points to the prefrontal cortex as an origin region of DS, while subcortical, occipital, and temporal regions are also implicated. The present study aimed to examine the contribution of communications between these brain regions to visual DS. To do it, we recruited two independent cohorts of participants for the study. One cohort participated in a visual search experiment where a salient distractor triggering distractor suppression to measure their DS and the other cohort filled out a Cognitive Failure Questionnaire to assess distractibility in daily life. Both cohorts collected resting-state functional magnetic resonance imaging (rs-fMRI) data to investigate function connectivity (FC) underlying DS. First, we generated predictive models of the DS measured in visual search task using resting-state functional connectivity between large anatomical regions. It turned out that the models could successfully predict individual's DS, indicated by a significant correlation between the actual and predicted DS (r = 0.32, p < 0.01). Importantly, Prefrontal-Temporal, Insula-Limbic and Parietal-Occipital connections contributed to the prediction model. Furthermore, the model could also predict individual's daily distractibility in the other independent cohort (r = –0.34, p < 0.05). Our findings showed the efficiency of the predictive models of distractor suppression encompassing connections between large anatomical regions and highlighted the importance of the communications between attention-related and visual information processing regions in distractor suppression. Current findings may potentially provide neurobiological markers of visual distractor suppression. [ABSTRACT FROM AUTHOR]