Correct Prediction Research Articles

An assessment of heterogenous ensemble classifiers for analyzing change‐proneness in open‐source software systems

AbstractSoftware managers constantly look out for methods that ensure cost effective development of good quality software products. An important means of accomplishing this is by allocating more resources to weak classes of a software product, which are prone to changes. Therefore, correct prediction of these change‐prone classes is critical. Though various researchers have investigated the performance of several algorithms for identifying them, the search for an optimum classifier still persists. To this end, this study critically investigates the use of six Heterogenous Ensemble Classifiers (HEC) for Software Change Prediction (SCP) by empirically validating datasets obtained from 12 open‐source software systems. The results of the study are statistically assessed using three robust performance indicators (AUC, F‐measure and Mathew Correlation Coefficient) in two different validation scenarios (within project and cross‐project). They indicate the superiority of Average Probability Voting Ensemble, a heterogenous classifier for determining change‐proneness in the investigated systems. The average AUC values of software change prediction models developed using this ensemble classifier exhibited an improvement of 3%‐9% and 3%‐11% respectively when compared with its base learners and homogeneous counter parts. Similar observations were inferred using other investigated performance measures. Furthermore, the evidence obtained from the results suggests that the change in number of base learners or type of meta‐learner does not exhibit significant change in the performance of corresponding heterogenous ensemble classifiers.

Predictions of Boron Phase Stability Using an Efficient Bayesian Machine Learning Interatomic Potential.

Thermodynamic phase stability of three elemental boron allotropes, i.e., α-B, β-B, and γ-B, was investigated using a Bayesian interatomic potential trained via a sparse Gaussian process (SGP). SGP potentials trained with data sets from on-the-fly active learning achieve quantum mechanical level accuracy when employed in molecular dynamics (MD) simulations to predict wide-ranging thermodynamic, structural, and vibrational properties. The simulated phase diagram (500-1400 K and 0-16 GPa) agrees with experimental measurements. The SGP-based MD simulations also successfully predicted that the B13 defect is critical in stabilizing β-B below 700 K. At higher temperatures, the entropy becomes the dominant factor, making β-B the more stable phase over α-B. This letter demonstrates that SGP potentials based on a training set consisting of defect-free-only systems could make correct predictions of defect-related phenomena in solid-state crystals, paving the path to investigate crystal phase stability and transitions.

Utility of Platelet Indices as Prognostic Markers of Sepsis: A Medical Intensive Care Unit-Based Cross-Sectional Study at a Rural Setup.

Introduction Even after the breakthrough advancements in the management and prognostic scoring of sepsis, it remains an important cause of morbidity and mortality encountered in intensive care units (ICUs) throughout the globe. This study highlights the utility of platelet indices as prognostic markers of sepsis. Methods In the present prospective cross-sectional study, a total of 177 patients with sepsis were enrolled using the Sepsis-3 criteria. The platelet indices were then linked to severity using the Acute Physiology and Chronic Health Evaluation (APACHE) II score. The correlation of platelet indices to morbidity in terms of the length of ICU stay, need for a mechanical ventilator, types of infection, and mortality was also assessed. Results The results showed that mean platelet volume (p = 0.004) and platelet distribution width (PDW; p = 0.009) were positively correlated with the severity of sepsis. Among all the parameters, plateletcrit (%) was the best predictor of the need for an invasive mechanical ventilator at a cutoff point of ≤0.22 with a 60.90% chance of correctly predicting the need for an invasive mechanical ventilator, as was mortality at a cutoff point of ≤0.22 with a 67.30% chance of correct prediction. Among the platelet indices, only PDW showed a significant association with growth in culture because patients with growth had significantly higher PDW as compared to those who did not have growth (22.4 ± 4.47 vs 20.81 ± 4.29, p = 0.011). Conclusion The difference between the survivors and non-survivor groups was statistically significant for platelet indices, making them easily available, cost-effective, and useful prognostic markers for patients in septic shock. This will help in easy understanding and preventing its morbid complications, even at the primary care physician level.

A method for detecting gas applied in seed germination with varying concentration based on gas sensors

Gas sensor is the technique of gas detection and it is the first attempt at the application of seed germination. However, for seed germination environment, concentrations of gases vary, increasing detection difficulty because of the slow response of gas sensor. In this study, the ability of gas sensors to detect gas applied in seed germination with varying concentration was researched. Ethylene and CO were employed and four gas sensors were applied. A new method was proposed. Back Propagation Neural Network (BPNN) was introduced and their performances were compared. The results showed that the proposed method had better overall performance, whose correct prediction rate for gas concentration fixed samples was 91.1% and for gas concentration varying samples was 83.3%. Moreover, the data size needed was small and the computation process was rapid. The results indicated that the proposed method was effective for gas application in seed germination with varying concentration.

Transfer learning graph representations of molecules for pKa, 13C-NMR, and solubility

We explore transfer learning models from a pre-trained graph convolutional neural network representation of molecules, obtained from SchNet, to predict 13C-NMR, pKa, and log S solubility. SchNet learns a graph representation of a molecule by associating each atom with an “embedding vector” and interacts the atom-embeddings with each other by leveraging graph convolutional filters on their interatomic distances. We pre-trained SchNet on molecular energy and demonstrate that the pre-trained atomistic embeddings can then be used as a transferable representation for a wide array of properties. On the one hand, for atomic properties such as micro-pK1 and 13C-NMR, we investigate two models, one linear and one neural net, that input pre-trained atom-embeddings of a particular atom (e.g. carbon) and predict a local property (e.g., 13C-NMR). On the other hand, for molecular properties such as solubility, a size-extensive graph model is built using the embeddings of all atoms in the molecule as input. For all cases, qualitatively correct predictions are made with relatively little training data (<1000 training points), showcasing the ease with which pre-trained embeddings pick up on important chemical patterns. The proposed models successfully capture well-understood trends of pK1 and solubility. This study advances our understanding of current neural net graph representations and their capacity for transfer learning applications in chemistry.

Development and validation of a quick, automated, and reproducible ATR FT-IR spectroscopy machine-learning model for Klebsiella pneumoniae typing.

The reliability of Fourier-transform infrared (FT-IR) spectroscopy for Klebsiella pneumoniae typing and outbreak control has been previously assessed, but issues remain in standardization and reproducibility. We developed and validated a reproducible FT-IR with attenuated total reflectance (ATR) workflow for the identification of K. pneumoniae lineages. We used 293 isolates representing multidrug-resistant K. pneumoniae lineages causing outbreaks worldwide (2002-2021) to train a random forest classification (RF) model based on capsular (KL)-type discrimination. This model was validated with 280 contemporaneous isolates (2021-2022), using wzi sequencing and whole-genome sequencing as references. Repeatability and reproducibility were tested in different culture media and instruments throughout time. Our RF model allowed the classification of 33 capsular (KL)-types and up to 36 clinically relevant K. pneumoniae lineages based on the discrimination of specific KL- and O-type combinations. We obtained high rates of accuracy (89%), sensitivity (88%), and specificity (92%), including from cultures obtained directly from the clinical sample, allowing to obtain typing information the same day bacteria are identified. The workflow was reproducible in different instruments throughout time (>98% correct predictions). Direct colony application, spectral acquisition, and automated KL prediction through Clover MS Data analysis software allow a short time-to-result (5 min/isolate). We demonstrated that FT-IR ATR spectroscopy provides meaningful, reproducible, and accurate information at a very early stage (as soon as bacterial identification) to support infection control and public health surveillance. The high robustness together with automated and flexible workflows for data analysis provide opportunities to consolidate real-time applications at a global level. IMPORTANCE We created and validated an automated and simple workflow for the identification of clinically relevant Klebsiella pneumoniae lineages by FT-IR spectroscopy and machine-learning, a method that can be extremely useful to provide quick and reliable typing information to support real-time decisions of outbreak management and infection control. This method and workflow is of interest to support clinical microbiology diagnostics and to aid public health surveillance.

Modeling of flaxseed protein, oil content, linoleic acid, and lignan content prediction based on hyperspectral imaging.

Protein, oil content, linoleic acid, and lignan are several key indicators for evaluating the quality of flaxseed. In order to optimize the testing methods for flaxseed's nutritional quality and enhance the efficiency of screening high-quality flax germplasm resources, we selected 30 flaxseed species widely cultivated in Northwest China as the subjects of our study. Firstly, we gathered hyperspectral information regarding the seeds, along with data on protein, oil content, linoleic acid, and lignan, and utilized the SPXY algorithm to classify the sample set. Subsequently, the spectral data underwent seven distinct preprocessing methods, revealing that the PLSR model exhibited superior performance after being processed with the SG smoothing method. Feature wavelength extraction was carried out using the Successive Projections Algorithm (SPA) and the Competitive Adaptive Reweighted Sampling (CARS). Finally, four quantitative analysis models, namely Partial Least Squares Regression (PLSR), Support Vector Regression (SVR), Multiple Linear Regression (MLR), and Principal Component Regression (PCR), were individually established. Experimental results demonstrated that among all the models for predicting protein content, the SG-CARS-MLR model predicted the best, with and of 0.9563 and 0.9336, with the corresponding Root Mean Square Error Correction (RMSEC) and Root Mean Square Error Prediction (RMSEP) of 0.4892 and 0.5616, respectively. In the optimal prediction models for oil content, linoleic acid and lignan, the was 0.8565, 0.8028, 0.9343, and the RMSEP was 0.8682, 0.5404, 0.5384, respectively. The study results show that hyperspectral imaging technology has excellent potential for application in the detection of quality characteristics of flaxseed and provides a new option for the future non-destructive testing of the nutritional quality of flaxseed.

Evaluating the Impact of Using Multiple-Metal Layers on the Layout Area of Switch Blocks for Tile-Based FPGAs in FinFET 7nm

A new area model for estimating the layout area of switch blocks is introduced in this work. The model is based on a realistic layout strategy. As a result, it not only takes into consideration the active area that is needed to construct a switch block but also the number of metal layers available and the actual dimensions of these metals. The model assigns metal layers to the routing tracks in a way that reduces the number of vias that are needed to connect different routing tracks together while maintaining the tile-based structure of FPGAs. It also takes into account the wiring area required for buffer insertion for long wire segments. The model is evaluated based on the layouts constructed in the ASAP7 FinFET 7nm Predictive Design Kit. We found that the new model, while specific to the layout strategy that it employs, improves upon the traditional active-based area estimation models by considering the growth of the metal area independently from the growth of the active area. As a result, the new model is able to more accurately estimate the layout area by predicting when the metal area will overtake the active area as the number of routing tracks is increased. This ability allows the more accurate estimation of the true layout cost of FPGA fabrics at the early floor planning and architectural exploration stage; and this increase in accuracy can encourage a wider use of custom FPGA fabrics that target specific sets of benchmarks in future SOC designs. Furthermore, our data indicate that the conclusions drawn from several significant prior architectural studies remain to be correct under FinFET geometries and wiring area considerations despite their exclusive use of active-only area models. This correctness is due to the small channel widths, around 30–60 tracks per channel, of the architectures that these studies investigate. For architectures that approach the channel width of modern commercial FPGAs with more than 100–200 tracks per channel, our data show that wiring area models justified by detailed layout considerations are an essential addition to active area models in the correct prediction of the implementation area of FPGAs.

Linking Gulf Stream air–sea interactions to the exceptional blocking episode in February 2019: a Lagrangian perspective

Abstract. The development of atmospheric blocks over the North Atlantic–European region can lead to extreme weather events like heat waves or cold air outbreaks. Despite their potential severe impact on surface weather, the correct prediction of blocking lifecycles remains a key challenge in current numerical weather prediction (NWP) models. Increasing evidence suggests that latent heat release in cyclones, the advection of cold air (cold air outbreaks, CAOs) from the Arctic over the North Atlantic, and associated air–sea interactions over the Gulf Stream are key processes contributing to the onset, maintenance, and persistence of such flow regimes. To better understand the mechanism connecting air–sea interactions over the Gulf Stream with changes in the large-scale flow, we focus on an episode between 20 and 27 February 2019, when a quasi-stationary upper-level ridge was established over western Europe accompanied by an intensified storm track in the northwestern North Atlantic. During that time, a record-breaking winter warm spell occurred over western Europe bringing temperatures above 20 ∘C to the United Kingdom, the Netherlands, and northern France. The event was preceded and accompanied by the development of several rapidly intensifying cyclones that originated in the Gulf Stream region and traversed the North Atlantic. To explore the mechanistic linkage between the formation of this block and air–sea interactions over the Gulf Stream, we adopt a Lagrangian perspective, using kinematic trajectories. This allows us to study the pathways and transformations of air masses that form the upper-level potential vorticity anomaly and interact with the ocean front. We establish that more than one-fifth of these air masses interact with the Gulf Stream in the lower troposphere, experiencing intense heating and moistening over the region due to the frequent occurrence of CAOs behind the cold front of the cyclones. Trajectories moistened by the advection of cold air over a warm ocean by one cyclone later ascend into the upper troposphere with the ascending airstream of a subsequent cyclone, fueled by the strong surface fluxes. These findings highlight the importance of CAOs in the Gulf Stream region, indicating that their intense coupling between the ocean and atmosphere plays a role in block development. Additionally, they provide a mechanistic pathway linking air–sea interactions in the lower troposphere and the upper-level flow.

Galaxy stellar and total mass estimation using machine learning

ABSTRACT Conventional galaxy mass estimation methods suffer from model assumptions and degeneracies. Machine learning (ML), which reduces the reliance on such assumptions, can be used to determine how well present-day observations can yield predictions for the distributions of stellar and dark matter. In this work, we use a general sample of galaxies from the TNG100 simulation to investigate the ability of multibranch convolutional neural network (CNN) based ML methods to predict the central (i.e. within 1−2 effective radii) stellar and total masses, and the stellar mass-to-light ratio (M*/L). These models take galaxy images and spatially resolved mean velocity and velocity dispersion maps as inputs. Such CNN-based models can, in general, break the degeneracy between baryonic and dark matter in the sense that the model can make reliable predictions on the individual contributions of each component. For example, with r-band images and two galaxy kinematic maps as inputs, our model predicting M*/L has a prediction uncertainty of 0.04 dex. Moreover, to investigate which (global) features significantly contribute to the correct predictions of the properties above, we utilize a gradient-boosting machine. We find that galaxy luminosity dominates the prediction of all masses in the central regions, with stellar velocity dispersion coming next. We also investigate the main contributing features when predicting stellar and dark matter mass fractions (f*, fDM) and the dark matter mass MDM, and discuss the underlying astrophysics.

Process-based classification of Mediterranean cyclones using potential vorticity

Abstract. Mediterranean cyclones (MCs) govern extreme weather events across the Euro-African Basin, affecting the lives of hundreds of millions. Despite many studies addressing MCs in the last few decades, their correct simulation and prediction remain a significant challenge to the present day, which may be attributed to the large variability among MCs. Past classifications of MCs are primarily based on geographical and/or seasonal separations; however, here we focus on cyclone genesis and deepening mechanisms. A variety of processes combine to govern MC genesis and evolution, including adiabatic and diabatic processes, topographic influences, land–sea contrasts, and local temperature anomalies. As each process bears a distinct signature on the potential vorticity (PV) field, a PV approach is used to distinguish among different “types” of MCs. Here, a combined cyclone-tracking algorithm is used to detect 3190 Mediterranean cyclone tracks in ECMWF ERA5 from 1979–2020. Cyclone-centered, upper-level isentropic PV structures in the peak time of each cyclone track are classified using a self-organizing map (SOM). The SOM analysis reveals nine classes of Mediterranean cyclones, with distinct Rossby-wave-breaking patterns, discernible in corresponding PV structures. Although classified by upper-level PV structures, each class shows different contributions of lower-tropospheric PV and flow structures down to the surface. Unique cyclone life cycle characteristics, associated hazards (precipitation, winds, and temperature anomalies), and long-term trends, as well as synoptic, thermal, dynamical, seasonal, and geographical features of each cyclone class, indicate dominant processes in their evolution. Among others, the classification reveals the importance of topographically induced Rossby wave breaking to the generation of the most extreme Mediterranean cyclones. These results enhance our understanding of MC predictability by linking the large-scale Rossby wave formations and life cycles to coherent classes of under-predicted cyclone aspects.

Abstract WP236: Harmonization of Stroke Risk Prediction Variables Using Natural Language Processing

Introduction: Early identification of stroke risk can profoundly influence an individual's chance of survival. While machine learning-based stroke risk prediction models perform better than traditional counterparts, they require integrating diverse open data sources. Variable discrepancies - where the same concept is described differently - pose challenges for data integration, requiring time-intensive manual harmonization. To address this, we developed an automated harmonization technique using natural language models. Methods: We utilized data from the Atherosclerosis Risk in Communities (ARIC), Multi-Ethnic Study of Atherosclerosis (MESA), Framingham Offspring Study, and Reasons for Geographic and Racial Differences in Stroke (REGARDS) cohorts. Our approach involved training a two-layer fully connected neural network (FCN) to predict a harmonized ‘concept variable’ based on variable descriptions. These predictions employed embeddings from Bidirectional Encoder Representations from Transformers for Biomedical Text Mining (BioBERT). Further refining was achieved through paired sentence classification tasks to determine if description pairs shared identical concepts. We experimented with both separate cohort training and combined cohort training to evaluate context dependency. We used cosine similarity between the embeddings of the pre-trained BioBERT encoder as a prediction score for our baseline method. Results: The harmonization methods using natural language models outperformed the baseline method. The FCN model improved the area under the receiver operating characteristic from 0.787 (baseline) to 0.985 for the paired classification task. For harmonization concept prediction, the top-2 accuracy - correct prediction in the top two answers - improved from 11.9% (baseline) to 54.6%, while the top-5 accuracy improved from 16.1% (baseline) to 67.6%. Conclusions: Utilizing natural language processing for data harmonization provides a scalable approach that improves accuracy and efficiency of prediction. This facilitates the inclusion of diverse cohorts, broadening the sample size and range of indicators available, which in turn advances the versatility and the generalizability of stroke risk prediction models.

Unplanned tooth movement in deepbite correction with Invisalign: A retrospective study

This study aimed to investigate unplanned tooth movements in deepbite correction with Invisalign. The sample consisted of 34 adult patients with deepbite malocclusion treated with Invisalign. Pretreatment and predicted digital models were exported from ClinCheck software, while the post-treatment model was obtained from an intraoral scan taken at the end of the first set of aligners. Digital models' superimposition was done using the eModel "Compare" software. Frequency and percentage of occurrence of unplanned linear and angular movements, their magnitude, and direction were then determined. In addition, comparison of unplanned movements was performed between anterior and posterior teeth as well as between linear and angular measurements. The percentage of unplanned movements ranged from 1.68% to 25.63% in the maxilla and 3.36% to 23.95% in the mandible, being most evident in the first and second molars. Unplanned movements were statistically significantly more frequent in posterior teeth compared with the anterior teeth (P > 0.05). In addition, the frequency of angular movements was statistically significantly higher for each tooth than linear movement (P < 0.05) with a higher mean value. Unplanned mesiodistal movements were the least frequent. Unplanned linear as well as angular movements were evident in almost all teeth. Unplanned faciolingual inclination and mesiodistal angulation movements were the most frequently observed. Unplanned vertical movements can affect the predictability of deepbite correction and thus merit particular attention in those cases.

Recent Progresses on Pathophysiology, Diagnosis, Therapeutic Modalities, and Management of Non-alcoholic Fatty Liver Disorder

Abstract: Non-alcoholic fatty liver disease (NAFLD) is currently the utmost common chronic liver disorder that happens through all age groups and is identified to occur in 14%-30% of the general population, demonstrating a critical and grossing clinical issue because of the growing incidence of obesity and overweight. From the histological aspect, it looks like alcoholic liver damage, but it happens in patients who avoid remarkable alcohol usage. NAFLD comprises a broad spectrum, ranging from benign hepatocellular steatosis to inflammatory nonalcoholic steatohepatitis (NASH), different levels of fibrosis, and cirrhosis. Patients with NASH are more susceptible to more rapid progression to cirrhosis and hepatocellular carcinoma. There is no single factor that drives proceeding from simple steatosis to NASH. However, a combination of multi parameters such as genetic background, gut microflora, intake of high fat/ fructose dietary contents or methionine/choline-deficient diet, and consequently accumulated hepatocellular lipids mainly including triglycerides and also other bio-analytes, such as free fatty acids, cholesterol, and phospholipids display a crucial role in disease promotion. NAFLD is related to overweight and insulin resistance (IR) and is regarded as the hepatic presentation of the metabolic syndrome, an amalgamation of medical statuses such as hyperlipidemia, hypertension, type 2 diabetes, and visceral obesity. Despite the increasing prevalence of this disease, which imposes a remarkable clinical burden, most affected patients remain undiagnosed in a timely manner, largely related to the asymptomatic entity of NAFLD patients and the unavailability of accurate and efficient noninvasive diagnostic tests. However, liver biopsy is considered a gold standard for NAFLD diagnosis, but due to being expensive and invasiveness is inappropriate for periodic disease screening. Some noninvasive monitoring approaches have been established recently for NAFLD assessment. In addition to the problem of correct disease course prediction, no effective therapeutic modalities are approved for disease treatment. Imaging techniques can commonly validate the screening and discrimination of NAFLD; nevertheless, staging the disease needs a liver biopsy. The present therapeutic approaches depend on weight loss, sports activities, and dietary modifications, although different insulin-sensitizing drugs, antioxidants, and therapeutic agents seem hopeful. This review aims to focus on the current knowledge concerning epidemiology, pathogenesis, and different biochemical experiments and imaging modalities applied to diagnose the different grades of NAFLD and its management, as well as new data about pharmacological therapies for this disorder.

Scheimpflug-Based Corneal Biomechanical Analysis As A Predictor of Glaucoma in Eyes With High Myopia.

To address if corneal biomechanical behavior has a predictive value for the presence of glaucomatous optical neuropathy in eyes with high myopia. This observational cross-sectional study included 209 eyes from 108 consecutive patients, divided into four groups: high myopia and primary open-angle glaucoma (POAG) - HMG, n = 53; high myopia without POAG - HMNG, n = 53; non-myopic with POAG - POAG, n = 50; non-myopic and non-POAG- NMNG, n = 53. Biomechanical assessment was made through a Scheimpflug-camera-based technology. Receiver operating characteristic curves were made for the discrimination between groups. Multivariable logistic regression models were performed to address the predictive value of corneal biomechanics for the presence of glaucoma. Areas Under the Receiver Operating Characteristic (AUROCs) above 0.6 were found in 6 parameters applied to discriminate between HMG and HMNG and six parameters to discriminate between POAG and NMNG. The biomechanical models with the highest power of prediction for the presence of glaucoma included 5 parameters with an AUROC of 0.947 for eyes with high myopia and 6 parameters with an AUROC of 0.857 for non-myopic eyes. In the final model, including all eyes, and adjusted for the presence of high myopia, the highest power of prediction for the presence of glaucoma was achieved including eight biomechanical parameters, with an AUROC of 0.917. Corneal biomechanics demonstrated differences in eyes with glaucoma and mainly in myopic eyes. A biomechanical model based on multivariable logistic regression analysis and adjusted for high myopia was built, with an overall probability of 91.7% for the correct prediction of glaucomatous damage.

ПРОГНОСТИЧЕСКАЯ ЗНАЧИМОСТЬ ИССЛЕДОВАНИЯ УРОВНЕЙ МАРКЕРОВ АПОПТОЗА – БЕЛКОВ P53 И BCL-2 У ПАЦИЕНТОВ ПОЖИЛОГО ВОЗРАСТА С ВИРУСНОЙ ПНЕВМОНИЕЙ, ВЫЗВАННОЙ SARS-COV-2

Abstract. Introduction. Тo date, the prognostic value has been proven for some laboratory biomarkers and biomarker panels for risk stratification in patients with viral pneumonia caused by SARS-CoV-2. In studies related to the research in the leading pathogenetic mechanisms of COVID-19, much attention is paid to apoptosis and its markers, such as proteins P53 and BCL-2. Aim: To study the levels of P53 and Bcl-2 proteins and their prognostic significance in lethal outcomes among elderly patients with viral pneumonia caused by SARS-CoV-2, hospitalized in an infectious hospital. Materials and Methods. For the period December 2021 – February 2022, in accordance with the inclusion criteria, a sample of 67 patients was formed from elderly patients (60-74 years old) hospitalized in the hospital of infectious diseases. During the follow-up, 18 patients (26.9 %) had fatal outcomes and 49 patients (73.1 %) had favorable outcomes. Biochemical markers of apoptosis, P53 and BCL-2, were detected using the enzyme immunoassay method. Results and Discussion. The analysis showed the presence of statistically significant differences when comparing the levels of both proteins, P53 (p=0.001) and BCL-2 (p=0.002) in patients with favorable outcomes, compared to those with fatal outcomes. Threshold values of the P53 and BCL-2 proteins at the cut-off point in case of death of elderly patients with viral pneumonia caused by SARS-CoV-2 were 99.3 pg/ml and 78.9 IU/ml, respectively. The classification tree includes the following lethality predictors: Threshold levels of apoptosis markers, the number of comorbid diseases, and the respiratory failure degree. Five terminal nodes were identified. Terminal nodes 6 and 8 were classified as high risk of death. Lethal outcomes were observed in 75.0 % of patients at these nodes. At terminal node 3, fatal outcome was observed in 8.1 % of patients, while there were no fatal outcomes at terminal node 7; therefore, both were classified as low-risk ones in terms of a fatal outcome. The sensitivity of the resulting model was 66.7 %, while its specificity was 91.8%. The total share of correct predictions made 85.1% among the subjects of this sample (p=0.044). Conclusions. The data obtained indicate the prognostic significance of studying the apoptosis markers, i. e., proteins P53 and BCL-2, in elderly patients with viral pneumonia caused by SARS-CoV-2. The classification tree proposed can be used as an auxiliary algorithm to optimize the personalized stratification of the lethality risk in elderly patients with viral pneumonia caused by SARS-CoV-2.

Comparative evaluation of bioinformatic tools for virus-host prediction and their application to a highly diverse community in the Cuatro Ciénegas Basin, Mexico.

Due to the enormous diversity of non-culturable viruses, new viruses must be characterized using culture-independent techniques. The associated host is an important phenotypic feature that can be inferred from metagenomic viral contigs thanks to the development of several bioinformatic tools. Here, we compare the performance of recently developed virus-host prediction tools on a dataset of 1,046 virus-host pairs and then apply the best-performing tools to a metagenomic dataset derived from a highly diverse transiently hypersaline site known as the Archaean Domes (AD) within the Cuatro Ciénegas Basin, Coahuila, Mexico. Among host-dependent methods, alignment-based approaches had a precision of 66.07% and a sensitivity of 24.76%, while alignment-free methods had an average precision of 75.7% and a sensitivity of 57.5%. RaFAH, a virus-dependent alignment-based tool, had the best overall performance (F1_score = 95.7%). However, when predicting the host of AD viruses, methods based on public reference databases (such as RaFAH) showed lower inter-method agreement than host-dependent methods run against custom databases constructed from prokaryotes inhabiting AD. Methods based on custom databases also showed the greatest agreement between the source environment and the predicted host taxonomy, habitat, lifestyle, or metabolism. This highlights the value of including custom data when predicting hosts on a highly diverse metagenomic dataset, and suggests that using a combination of methods and qualitative validations related to the source environment and predicted host biology can increase the number of correct predictions. Finally, these predictions suggest that AD viruses infect halophilic archaea as well as a variety of bacteria that may be halophilic, halotolerant, alkaliphilic, thermophilic, oligotrophic, sulfate-reducing, or marine, which is consistent with the specific environment and the known geological and biological evolution of the Cuatro Ciénegas Basin and its microorganisms.

Corrective Day-ahead Prediction of Solar Radiation based on Online Measurements

In recent years, reduction of greenhouse gas emissions has been a serious issue throughout the world. The Japanese Government aims to reduce 25% of the emissions in 1990 by 2020. Eco-cities which effectively use renewable energy sources attract an attention in order to attain this goal. A smart energy town project in Honjyo, Japan has been conducted by making use of renewable energy sources. Above all, photovoltaic (PV) generators are expected to play a key role in this project. In addition, PV generators are expected to be installed in the residential and commercial areas due to ease of its installation. However, unstable power output from PV generation causes negative impacts on power quality of distribution networks. Therefore, prediction of PV output is indispensable to operate networks properly. In this paper, we propose a method to predict a day-ahead solar radiation and output power of PV generation in specific areas. Furthermore, the predicted solar radiation is corrected using the on-line measured actual data. As a result, the correction of a day-ahead prediction proposed here reduces errors of daily solar radiation by around 34% .

Key Grazing Behaviours of Beef Cattle Identify Specific Genotypes of the Glutamate Metabotropic Receptor 5 Gene (GRM5)

Genotype-phenotype associations between the bovine genome and grazing behaviours measured over time and across contexts have been reported in the past decade, with these suggesting the potential for genetic control over grazing personalities in beef cattle. From the large array of metrics used to describe grazing personality behaviours (GP-behaviours), it is still unclear which ones are linked to specific genes. Our prior observational study has reported associations and trends towards associations between genotypes of the glutamate metabotropic receptor 5 gene (GRM5) and four GP-behaviours, yet the unbalanced representation of GRM5 genotypes occurring in observational studies may have limited the ability to detect associations. Here, we applied a subsampling technique to create a genotypically-balanced dataset in a quasi-manipulative experiment with free ranging cows grazing in steep and rugged terrain of New Zealand’s South Island. Using quadratic discriminant analysis, two combinations of eleven GP-behaviours (and a total of fifteen behaviours) were selected to build an exploration model and an elevation model, respectively. Both models achieved ∼ 86% accuracy in correctly discriminating cows’ GRM5 genotypes with the training dataset, and the exploration model achieved 85% correct genotype prediction of cows from a testing dataset. Our study suggests a potential pleiotropic effect, with GRM5 controlling multiple grazing behaviours, and with implications for the grazing of steep and rugged grasslands. The study highlights the importance of grazing behavioural genetics in cattle and the potential use of GRM5 markers to select individuals with desired grazing personalities and built herds that collectively utilize steep and rugged rangelands sustainably.

Using simulated microhaplotype genotyping data to evaluate the value of machine learning algorithms for inferring DNA mixture contributor numbers

Inferring the number of contributors (NoC) is a crucial step in interpreting DNA mixtures, as it directly affects the accuracy of the likelihood ratio calculation and the assessment of evidence strength. However, obtaining the correct NoC in complex DNA mixtures remains challenging due to the high degree of allele sharing and dropout. This study aimed to analyze the impact of allele sharing and dropout on NoC inference in complex DNA mixtures when using microhaplotypes (MH). The effectiveness and value of highly polymorphic MH for NoC inference in complex DNA mixtures were evaluated through comparing the performance of three NoC inference methods, including maximum allele count (MAC) method, maximum likelihood estimation (MLE) method, and random forest classification (RFC) algorithm. In this study, we selected the top 100 most polymorphic MH from the Southern Han Chinese (CHS) population, and simulated over 40 million complex DNA mixture profiles with the NoC ranging from 2 to 8. These profiles involve unrelated individuals (RM type) and related pairs of individuals, including parent-offspring pairs (PO type), full-sibling pairs (FS type), and second-degree kinship pairs (SE type). Our results indicated that how the number of detected alleles in DNA mixture profiles varied with the markers’ polymorphism, kinship’s involvement, NoC, and dropout settings. Across different types of DNA mixtures, the MAC and MLE methods performed best in the RM type, followed by SE, FS, and PO types, while RFC models showed the best performance in the PO type, followed by RM, SE, and FS types. The recall of all three methods for NoC inference were decreased as the NoC and dropout levels increased. Furthermore, the MLE method performed better at low NoC, whereas RFC models excelled at high NoC and/or high dropout levels, regardless of the availability of a priori information about related pairs of individuals in DNA mixtures. However, the RFC models which considered the aforementioned priori information and were trained specifically on each type of DNA mixture profiles, outperformed RFC_ALL model that did not consider such information. Finally, we provided recommendations for model building when applying machine learning algorithms to NoC inference.