Estimation Of Prediction Accuracy Research Articles

Estimation of genetic correlations and genomic prediction accuracy for reproductive and carcass traits in Hanwoo cows.

This study estimated the heritabilities (h2) and genetic and phenotypic correlations between reproductive traits, including calving interval (CI), age at first calving (AFC), gestation length (GL), number of artificial inseminations per conception (NAIPC), and carcass traits, including carcass weight (CWT), eye muscle area (EMA), backfat thickness (BF), and marbling score (MS) in Korean Hanwoo cows. In addition, the accuracy of genomic predictions of breeding values was evaluated by applying the genomic best linear unbiased prediction (GBLUP) and the weighted GBLUP (WGBLUP) method. The phenotypic data for reproductive and carcass traits were collected from 1,544 Hanwoo cows, and all animals were genotyped using Illumina Bovine 50K single nucleotide polymorphism (SNP) chip. The genetic parameters were estimated using a multi-trait animal model using the MTG2 program. The estimated h2 for CI, AFC, GL, NAIPC, CWT, EMA, BF, and MS were 0.10, 0.13, 0.17, 0.11, 0.37, 0.35, 0.27, and 0.45, respectively, according to the GBLUP model. The GBLUP accuracy estimates ranged from 0.51 to 0.74, while the WGBLUP accuracy estimates for the traits under study ranged from 0.51 to 0.79. Strong and favorable genetic correlations were observed between GL and NAIPC (0.61), CWT and EMA (0.60), NAIPC and CWT (0.49), AFC and CWT (0.48), CI and GL (0.36), BF and MS (0.35), NAIPC and EMA (0.35), CI and BF (0.30), EMA and MS (0.28), CI and AFC (0.26), AFC and EMA (0.24), and AFC and BF (0.21). The present study identified low to moderate positive genetic correlations between reproductive and CWT traits, suggesting that a heavier body weight may lead to a longer CI, AFC, GL, and NAIPC. The moderately positive genetic correlation between CWT and AFC, and NAIPC, with a phenotypic correlation of nearly zero, suggesting that the genotype-environment interactions are more likely to be responsible for the phenotypic manifestation of these traits. As a result, the inclusion of these traits by breeders as selection criteria may present a good opportunity for developing a selection index to increase the response to the selection and identification of candidate animals, which can result in significantly increased profitability of production systems.

A meta-analysis on the effects of marker coverage, status number, and size of training set on predictive accuracy and heritability estimates from genomic selection in tree breeding

Genomic selection (GS) is increasingly used in tree breeding because of the possibility to hasten breeding cycles, increase selection intensity or facilitate multi-trait selection, and to obtain less biased estimates of quantitative genetic parameters such as heritability. However, tree breeders are aiming to obtain accurate estimates of such parameters and breeding values while optimizing sampling and genotyping costs. We conducted a metadata analysis of results from 28 GS studies totalling 115 study-traits. We found that heritability estimates obtained using DNA marker-based information for a variety of traits and species were not significantly related to variation in the total number of markers ranging from about 1500 to 116 000, nor by the marker density, ranging from about 1 to 60 markers/centimorgan, nor by the status number of the breeding populations ranging from about 10 to 620, nor by the size of the training set ranging from 236 to 2458. However, the predictive accuracy of breeding values was generally higher when the status number of the breeding population was smaller, which was expected given the higher level of relatedness in small breeding populations, and the increased ability of a given number of markers to trace the long-range linkage disequilibrium in such conditions. According to expectations, the predictive accuracy also increased with the size of the training set used to build marker-based models. Genotyping arrays with a few to many thousand markers exist for several tree species and with the actual costs, GS could thus be efficiently implemented in many more tree breeding programs, delivering less biased genetic parameters and more accurate estimates of breeding values.

Harnessing the power of hybrid deep learning algorithm for the estimation of global horizontal irradiance

Accurately and precisely estimating global horizontal irradiance (GHI) poses significant challenges due to the unpredictable nature of climate parameters and geographical limitations. To address this challenge, this study proposes a forecasting framework using an integrated model of the convolutional neural network (CNN), long short-term memory (LSTM), and gated recurrent unit (GRU). The proposed model uses a dataset of four different districts in Rajasthan, each with unique solar irradiance patterns. Firstly, the data was preprocessed and then trained with the optimized parameters of the standalone and hybrid models and compared. It can be observed that the proposed hybrid model (CNN-LSTM-GRU) consistently outperformed all other models regarding Mean absolute error (MAE) and Root mean squared error (RMSE). The experimental results demonstrate that the proposed method forecasts accurate GHI with a RMSE of 0.00731, 0.00730, 0.00775, 0.00810 and MAE of 0.00516, 0.00524, 0.00552, 0.00592 for Barmer, Jaisalmer, Jodhpur and Bikaner respectively. This indicates that the model is better at minimizing prediction errors and providing more accurate GHI estimates. Additionally, the proposed model achieved a higher coefficient of determination (R (Ghimire et al., 2019)), suggesting that it best fits the dataset. A higher R2 value signifies that the proposed model could explain a significant portion of the variance in the GHI dataset, further emphasizing its predictive capabilities. In conclusion, this work demonstrates the effectiveness of the hybrid algorithm in improving adaptability and enhancing prediction accuracy for GHI estimation.

Accuracy of genomic prediction using multiple Atlantic salmon populations

BackgroundThe accuracy of genomic prediction is partly determined by the size of the reference population. In Atlantic salmon breeding programs, four parallel populations often exist, thus offering the opportunity to increase the size of the reference set by combining these populations. By allowing a reduction in the number of records per population, multi-population prediction can potentially reduce cost and welfare issues related to the recording of traits, particularly for diseases. In this study, we evaluated the accuracy of multi- and across-population prediction of breeding values for resistance to amoebic gill disease (AGD) using all single nucleotide polymorphisms (SNPs) on a 55K chip or a selected subset of SNPs based on the signs of allele substitution effect estimates across populations, using both linear and nonlinear genomic prediction (GP) models in Atlantic salmon populations. In addition, we investigated genetic distance, genetic correlation estimated based on genomic relationships, and persistency of linkage disequilibrium (LD) phase across these populations.ResultsThe genetic distance between populations ranged from 0.03 to 0.07, while the genetic correlation ranged from 0.19 to 0.99. Nonetheless, compared to within-population prediction, there was limited or no impact of combining populations for multi-population prediction across the various models used or when using the selected subset of SNPs. The estimates of across-population prediction accuracy were low and to some extent proportional to the genetic correlation estimates. The persistency of LD phase between adjacent markers across populations using all SNP data ranged from 0.51 to 0.65, indicating that LD is poorly conserved across the studied populations.ConclusionsOur results show that a high genetic correlation and a high genetic relationship between populations do not guarantee a higher prediction accuracy from multi-population genomic prediction in Atlantic salmon.

A hybrid network of NARX and DS-attention applied for the state estimation of lithium-ion batteries

Monitoring the state of the battery, including the state of charge (SOC) and state of health (SOH), is crucial for ensuring the safety and reliability of electrical equipment. The paper presents a novel hybrid network that combines nonlinear autoregressive model with exogenous inputs (NARX) and DS-attention. The proposed DS-attention method establishes a robust mapping relationship between inputs and outputs, it is a specialized method of the recurrent neural network that enhances the estimation performance by incorporating division function and self-adaptive function into the attention mechanism. The division function is designed to efficiently differentiate between exogenous inputs and state outputs, thereby minimizing the potential for cross-interference between them, and the self-adaptive function optimizes the query features within the attention mechanism. The results demonstrate that the hybrid method of NARX and DS-attention achieves higher prediction accuracy for SOH and SOC estimation of lithium-ion batteries. Specifically, compared with the best-performing similar methods, the proposed hybrid method enhanced the prediction accuracy by 22.9%, 60.7%, and 51.2% across three different SOH datasets, respectively. In terms of long sequence SOC forecasting, the improvements in prediction accuracy under two different working conditions are 82.5% and 60.5%, respectively. The proposed algorithm optimizes the attention mechanism based on the characteristics of the NARX, resulting in higher estimation accuracy for predicting the state of lithium batteries.

Optimizing feature selection and parameter tuning for breast cancer detection using hybrid GAHBA-DNN framework

Breast cancer has been life-threatening for many years as it is the common cause of fatality among women. The challenges of screening such tumors through manual approaches can be overcome by computer-aided diagnosis, which aids radiologists in making precise decisions. The selection of significant features is crucial for the estimation of prediction accuracy. This work proposes a hybrid Genetic Algorithm (GA) and Honey Badger Algorithm (HBA) based Deep Neural Network (DNN), HGAHBA-DNN for the concurrent optimal features selection and parameter optimization; further, the optimal features and parameters extracted are fed into the DNN for the prediction of the breast cancer. It fuses the benefits of HBA with parallel processing and efficient feedback with GA’s excellent global convergent rate during the processing stages. The aforementioned method is evaluated on the Wisconsin Original Breast Cancer (WOBC), Wisconsin Diagnostic Breast Cancer (WDBC), and the Surveillance, Epidemiology, and End Results (SEER) datasets. Subsequently, the performance is validated using several metrics like accuracy, precision, Recall, and F1-score. The experimental result shows that HGAHBA-DNN obtains accuracy of 99.42%, 99.84%, and 92.44% for the WOBC, WDBC, and SEER datasets respectively, which is much superior to the other state-of-the-art methods.

Genomic selection accelerates genetic improvement of resistance to Vibriosis in the Pacific oyster, Crassostrea gigas

Due to lack of acquired immune system, the oysters cultured along coasts are subject to frequent pathogen threats, which leads to severe disease outbreaks around the world. It's well recognized that the selection breeding of aquatic animals can be accelerated via the harnessing of genomic tools to increase genetic gain and shorten the breeding time. In this work, we carried out genomic selection breeding in the Pacific oysters (Crassostrea gigas) for genetic improvement of resistance to Vibriosis. The genome-wide variations were genotyped by ddRAD-seq from 295 oysters with contrasted resistance to Vibrio infection. Based on genome-wide SNPs, we performed an estimation of genomic heritability and prediction accuracy for resistance to Vibrio alginolyticus in C. gigas. The genomic heritability of resistance to V. alginolyticus was low to moderate, ranging from 0.1405 to 0.2730. Four genomic selection models including rrBLUP, Bayes A, Bayes B and Bayesian Lasso were evaluated, of which Bayes A showed superior prediction accuracy and computational speed. The genomic estimated breeding value (GEBV) calculated by genomic selection model can effectively distinguish the resistance or susceptibility of oysters to Vibriosis. Selection of individuals with high GEBV as broodstock greatly improved the resistance to Vibriosis of their progeny, resulting in 18.42% increase in relative survival rate and 12.73% increase in relative survival time compared to the control population. For the first time, this work reported the efficiency of genomic selection breeding for genetic improvement for resistance trait to Vibriosis in the C. gigas, which would greatly accelerate the cultivation of Vibriosis resistant oyster strains to support the healthy and sustainable development of aquaculture.

Methods and tools for assessing the accuracy of composite property models

Property modeling is essential for optimal application and efficient design of new composite materials. Although numerous models of composite properties are available in the literature, there is a lack of rigorous validation of these models. This paper introduces a freely distributed collection of experimental data on the electrical and dielectric properties of composites and an open-source program package for modeling and testing composite property models. The software provides a quantitative estimation of prediction accuracy, making it a valuable tool for assessing the quality of a composite model. The paper demonstrates the use of the proposed tools to estimate two-phase models of composite electrical conductivity that account for the filler shape. The results of this test showed that the examined models failed to simulate the shape of the filler. Also, the article gives an example of the comparison of three-phase models of relative permittivity of composites with fitting parameters. In addition, the study presents the analysis of errors in the approximation of experimental data by the model, as well as recommendations on methods and selection of criteria for evaluating the quality of models.

A computer code for optimizing the neutronics model parameters based on results of reactor physics experiments

The paper describes in brief the functional capabilities of a computer code for optimizing the neutronics model parameters (neutron data, technological parameters, and their covariance matrices) based on results of reactor physics experiments using conditional nonlinear multi-parameter optimization algorithms. The code’s application scope includes adjustment of neutron constants, technological parameters and their covariance matrices based on integral measurement results, formulation of requiremen117198ts with respect to the neutron data uncertainties for achieving the target accuracies in calculation of the reactor functionals, and estimation of the reactor performance prediction accuracy, as well as the informativity and similarity metrics of reactor physics experiments with respect to each other and in relation to the target reactor system. The paper also considers some examples of using the code to refine the neutronics models of nuclear reactor and fuel cycle systems based on results of reactor physics experiments.

Modeling user choice behavior under data corruption: Robust learning of the latent decision threshold model

Recent years have witnessed the emergence of many new mobile apps and user-centered systems that interact with users by offering choices with rewards. These applications have been promising to address challenging societal problems such as congestion in transportation and behavior changes for healthier lifestyles. Considerable research efforts have been devoted to model the user behaviors in these new applications. However, as real-world user data is often prone to data corruptions, the success of these models hinges on a robust learning method. Building on the recently proposed Latent Decision Threshold model, this article shows that, among the existing robust learning frameworks, the L 0-norm-based framework can outperform other state-of-the-art methods in terms of prediction accuracy and model estimation. And based on the L 0-norm framework, we further develop a user screening algorithm to identify potential bad actors.

PSVII-2 The Impact of Genomic Information Inclusion on Breeding Value Prediction for a Novel Parasite Resistance Indicator Trait in Brazilian Anglo-Nubian Goats: a Pilot Study

Abstract Infection by Haemonchus contortus is the gastrointestinal parasitosis of greatest economic and social importance in the production of goats, being the leading cause of mortality in the Brazilian goat industry. Genetic selection for resistance to parasites in Anglo-Nubian goats has proven to be a viable alternative to traditional methods of controlling gastrointestinal nematodes, producing animals genetically superior in resistance to parasites. This study aimed to evaluate the impact of including genomic information on the estimation of variance components, genetic parameters, and predictive accuracy of genetic values for the new trait Parasitic Resistance (PR) in Anglo-Nubian goats. The PR is generated by measuring Body Condition Score, FAMACHA chart (ranging from 1 to 5), and Fecal Egg Count on a regular scale. These three traits were used as input variables in the CAPRIOVI software, which uses fuzzy logic to create the PR score, ranging from 0 to 10. Animals with lower scores are more likely to be parasitized and require treatment, while animals with higher scores are less likely to be infected. We used information from 352 animals, totaling 3,855 repeated measurements of PR over the years (2001 to 2020) and 1,500 animals in the pedigree. Animals with less than three measurements for any of the three input traits were removed from the dataset. Contemporary groups were built from information from the year of birth, data collection season, birth season, and physiological stage of the animal, considered as fixed effects, and age at the time of data collection as a covariate in an animal repeatability model. A total of 89 animals were genotyped using the Bead Chip Goats SNP50. After quality control, the remaining 38,659 SNPs (Table 1) were used to build the hybrid matrix (H). The variance components (VC) were estimated using the Bayesian Gibbs Sampling method, and breeding values were predicted using the Best Linear Unbiased Prediction (BLUP) and single-step Best Linear Unbiased Prediction (ssGBLUP) methods. The additive genetic variance and heritability estimation ranged from 0.13 to 0.14 and 0.04 to 0.05, respectively (Table 2). The accuracies obtained for PR with BLUP and ssGBLUP were 0.12 and 0.14, respectively, representing a gain of 16.7%. The genomic information (25 % of all animals with phenotypes) with the repeated records improved the predictive model capability based on accuracy. It also improved the VC estimation for PR in Anglo-Nubian goats raised in the Brazilian semi-arid region. These results highlight the positive impact of genetic evaluation of goats for PR, even with few phenotypes and genotyped animals. In addition to elucidating the additive genetic influence of parasite resistance, it can be a great source of information on the selection and production of Anglo-Nubian goats.

Fractional Flow Reserve-Based Patient Risk Classification.

Cardiovascular diseases (CVDs) are a leading cause of death. If not treated in a timely manner, cardiovascular diseases can cause a plethora of major life complications that can include disability and a loss of the ability to work. Globally, acute myocardial infarction (AMI) is responsible for about 3 million deaths a year. The development of strategies for prevention, but also the early detection of cardiovascular risks, is of great importance. The fractional flow reserve (FFR) is a measurement used for an assessment of the severity of coronary artery stenosis. The goal of this research was to develop a technique that can be used for patient fractional flow reserve evaluation, as well as for the assessment of the risk of death via gathered demographic and clinical data. A classification ensemble model was built using the random forest machine learning algorithm for the purposes of risk prediction. Referent patient classes were identified by the observed fractional flow reserve value, where patients with an FFR higher than 0.8 were viewed as low risk, while those with an FFR lower than 0.8 were identified as high risk. The final classification ensemble achieved a 76.21% value of estimated prediction accuracy, thus achieving a mean prediction accuracy of 74.1%, 77.3%, 78.1% and 83.6% over the models tested with 5%, 10%, 15% and 20% of the test samples, respectively. Along with the machine learning approach, a numerical approach was implemented through a 3D reconstruction of the coronary arteries for the purposes of stenosis monitoring. Even with a small number of available data points, the proposed methodology achieved satisfying results. However, these results can be improved in the future through the introduction of additional data, which will, in turn, allow for the utilization of different machine learning algorithms.

Estimation of genetic correlations and genomic prediction accuracy for reproductive and carcass traits in Hanwoo cows

Estimation of genetic correlations and genomic prediction accuracy for reproductive and carcass traits in Hanwoo cows

Frailty and risk of adverse outcomes among community-dwelling older adults in China: a comparison of four different frailty scales.

Data on which frailty scales are most suitable for estimating risk in Chinese community populations remain limited. Herein we examined and compared four commonly used frailty scales in predicting adverse outcomes in a large population-based cohort of Chinese older adults. A total of 5402 subjects (mean age 66.3 ± 9.6 years, 46.6% male) from the WHO Study on global AGEing and adult health (SAGE) in Shanghai were studied. Frailty was measured using a 35-item frailty index (FI), the frailty phenotype (FP), FRAIL, and Tilburg Frailty Indicator (TFI). Multivariate logistic regression models were performed to evaluate the independent association between frailty and outcomes including 4-year disability, hospitalization, and 4- and 7-year all-cause mortality. The accuracy for predicting these outcomes was determined by evaluating the area under the curve (AUC). The prevalence of frailty, sensitivity, and specificity were calculated using our proposed cut-off points and other different values. Prevalence of frailty ranged from 4.2% (FRAIL) to 16.9% (FI). FI, FRAIL and TFI were comparably associated with 4-year hospitalization, and 4- and 7-year mortality (adjusted odds ratios [aORs] 1.44-1.69, 1.91-2.22 and 1.85-2.88, respectively). FRAIL conferred the greatest risk of 4-year disability, followed by FI and TFI (aOR 5.55, 3.50, and 1.91, respectively). FP only independently predicted 4- and 7-year mortality (aOR 1.57 and 2.21, respectively). AUC comparisons showed that FI, followed by TFI and FRAIL, exhibited acceptable predictive accuracy for 4-year disability, 4- and 7-year mortality (AUCs 0.76-0.78, 0.71-0.71, 0.65-0.72, respectively), whereas all scales poorly predicted 4-year hospitalization (AUCs 0.53-0.57). For each scale, while specificity estimates (85.3-97.3%) were high and similar across all outcomes, their sensitivity estimates (6.3-56.8%) were not sufficient yet. Prevalence of frailty, sensitivity, and specificity varied considerably when different cut-off points were used. Frailty defined using any of the four scales was associated with an increased risk of adverse outcomes. Although FI, FRAIL and TFI exhibited fair-to-moderate predictive accuracy and high specificity estimates, their sensitivity estimates were not sufficient yet. Overall, FI performed best in estimating risk, while TFI and FRAIL were additionally useful, the latter perhaps being more applicable to Chinese community-dwelling older adults.

A Probabilistic Architecture of Long-Term Vehicle Trajectory Prediction for Autonomous Driving

In mixed and dynamic traffic environments, accurate long-term trajectory forecasting of surrounding vehicles is one of the indispensable preconditions for autonomous vehicles (AVs) to accomplish reasonable behavioral decisions and guarantee driving safety. In this paper, we propose an integrated probabilistic architecture for long-term vehicle trajectory prediction, which consists of a driving inference model (DIM) and a trajectory prediction model (TPM). The DIM is designed and employed to accurately infer the potential driving intention based on a dynamic Bayesian network. The proposed DIM incorporates the basic traffic rules and multivariate vehicle motion information. To further improve the prediction accuracy and realize uncertainty estimation, we develop a Gaussian process (GP)-based TPM, considering both the short-term prediction results of the vehicle model and the driving motion characteristics. Afterward, the effectiveness of our novel approach is demonstrated by conducting experiments on a public naturalistic driving dataset under lane-changing scenarios. The superior performance on the task of long-term trajectory prediction is presented and verified by comparing with other advanced methods.

Degradation-Aware Ensemble of Diverse Predictors for Remaining Useful Life Prediction

Abstract A traditional ensemble approach to predicting the remaining useful life (RUL) of equipment and other assets has been constructing data-driven and model-based ensembles using identical predictors. This ensemble approach may perform well on quality data collected from laboratory tests but may ultimately fail when deployed in the field because of higher-than-expected noise, missing measurements, and different degradation trends. In such work environments, the high similarity of the predictors can lead to large under/overestimates of RUL, where the ensemble is only as accurate as the predictor which under/overestimated RUL the least. In response to this, we investigate whether an ensemble of diverse predictors might be able to predict RUL consistently and accurately by dynamically aggregating the predictions of various algorithms which are found to perform differently under the same conditions. We propose improving ensemble model performance by (1) using a combination of diverse learning algorithms which are found to perform differently under the same conditions and (2) training a data-driven model to adaptively estimate the prediction weight each predictor receives. The proposed methods are compared to three existing ensemble prognostics methods on open-source run-to-failure datasets from two popular systems of prognostics research: lithium-ion batteries and rolling element bearings. Results indicate the proposed ensemble method provides the most consistent prediction accuracy and uncertainty estimation quality across multiple test cases, whereas the individual predictors and ensembles of identical predictors tend to provide overconfident predictions.

A Predictive Scoring System for Therapy Failure in Persons with Chronic Myeloid Leukemia Receiving Initial a Second-Generation Tyrosine Kinase Inhibitor Therapy

Purposes Develop and validate a predictive scoring system for therapy failure in persons with chronic phase chronic myeloid leukemia (CML) receiving initial a second-generation tyrosine kinase inhibitor (2G-TKI). Methods Data from 329 consecutive subjects with CML receiving initial 2G-TKI therapy at one centre (Beijing, China) in a training dataset were interrogated to develop a predictive model which was then validated in 1,083 subjects from 65 other centers in China. Therapy-failure was defined using the 2020 ELN criteria including progression to accelerated or blast phases or CML-related death. Fine-Gray model was used for uni- and multi-variable analyses (UVA and MVA) to identify co-variates significantly-associated with therapy-failure. Akaike information criterion (AIC) was used to select prognostic co-variate and determine the best model. 1000 bootstrap re-samplings, Fine-Gray test, the minimal p-value approach, Bonferroni correction and kernel density estimator were used to determine optimal cut-offs to classify subjects into risk cohorts. Time-dependent receiver-operator characteristic curves (ROC) were used to estimate prediction accuracy and calibration plots used to determine concordance between predicted and observed cumulative incidences. Decision curve analysis (DCA) was used to calculate the net benefit using the model. Results In the training dataset, 279 subjects (85%) initially received nilotinib; and 50 (15%), dasatinib. Median 2G-TKI therapy-duration was 36 months (interquartile range [IQR], 24-68 months). 75 subjects (23%) had therapy-failure at a median of 9 months (IQR 3-12 months) after therapy start. In MVA, increasing splenomegaly and a higher percentage blood blasts were significantly-associated with the higher cumulative incidences of therapy-failure. We used these data to develop a predictive scoring system: "2G-TKI failure risk score= 0.6405×[(spleen size below the costal margin in cm + 0.2)/10] + 0.1348×(blood blasts + 0.2)" which divided subjects into low- (score ≤ 0.3021; n = 89; 31%), intermediate- (0.3021 < score < 1.4508; n = 155; 54%) and high-risk (score ≥ 1.4508; n = 43; 15%) cohorts with 5-year cumulative incidences of therapy-failure of 3% (95% Confidence Interval [CI], 0, 5%), 26% (20, 32%) and 72% (56, 88%; p-value for trend < 0.001; Figure 1A). Hazard ratios (HRs) for therapy-failure (low-risk cohort as reference) were 4.8 (1.7, 13.6; p = 0.002) and 16.5 (5.7, 47.6; p < 0.001; p for trend < 0.001) for the intermediate- and high-risk cohorts. In the external validation dataset, subjects were classified into the low- (n = 433; 40%), intermediate- (n = 583; 54%) and high-risk (n = 67; 6%) cohorts using our scoring system. 5-year cumulative incidences of therapy failure were 8% (3, 13%), 18% (12, 25%) and 44% (29, 60%; p for trend < 0.001; Figure 1B). HRs (low-risk cohort as reference) were 2.7 (1.9, 3.6; p < 0.001) and 10.6 (5.8, 15.4; p < 0.001; p for trend < 0.001) for the intermediate- and high-risk cohorts. In the training dataset cohort time-dependent AUROC curves for therapy-failure at 1-, 3-, and 5-years in the model were 0.73-0.82 and 0.68-0.75 in the validation dataset. Calibration plots for the 1-, 3- and 5-years cumulative incidences of therapy-failure indicated good concordance between predicted and observed outcomes. DCA curves indicated net benefit from using the predictive scoring system in clinical settings. Conclusions We developed and externally validated a therapy-failure scoring system in persons with chronic phase CML receiving initial 2G-TKI. The model had high AUROCs and performed well in calibration plots and DCAs. Our model may help physicians decide appropriateness of initial the 2G-TKI therapy including possible selection of appropriate haematopoietic cell transplant candidates. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal

Genetic prediction of male pattern baldness based on large independent datasets

Genetic prediction of male pattern baldness (MPB) is important in science and society. Previous genetic MPB prediction models were limited by sparse marker coverage, small sample size, and/or data dependency in the different analytical steps. Here, we present novel models for genetic prediction of MPB based on a large set of markers and large independent subsample sets drawn among 187,435 European subjects. We selected 117 SNP predictors within 85 distinct loci from a list of 270 previously MPB-associated SNPs in 55,573 males of the UK Biobank Study (UKBB). Based on these 117 SNPs with and without age as additional predictor, we trained, by use of different methods, prediction models in a non-overlapping subset of 104,694 UKBB males and tested them in a non-overlapping subset of 26,177 UKBB males. Estimates of prediction accuracy were similar between methods with AUC ranges of 0.725–0.728 for severe, 0.631–0.635 for moderate, 0.598–0.602 for slight, and 0.708–0.711 for no hair loss with age, and slightly lower without, while prediction of any versus no hair loss gave 0.690–0.711 with age and slightly lower without. External validation in an early-onset enriched MPB dataset from the Bonn Study (N = 991) showed improved prediction accuracy without considering age such as AUC of 0.830 for no vs. any hair loss. Because of the large number of markers and the large independent datasets used for the different analytical steps, the newly presented genetic prediction models are the most reliable ones currently available for MPB or any other human appearance trait.

A classification scheme to improve conclusion instability using Bellwether moving windows

AbstractContextThe use of a subset of recently completed and exemplary data, namely, Bellwether moving window (BMW) has proven successful to result in improved accuracy in software effort estimation (SEE). These outcomes were achieved based on the theory that estimation outcome of a future event depends on previous events. Thus, the existence of a BMW yield improved prediction accuracy for new project estimation. However, the conclusion instability problem across learners still threatens the reliability of SEE for new projects. Such instability concerns are attributed to the data subset considered for the training and validation needs of learners.ObjectiveTo investigate whether the use of BMWs together with an effort classification scheme can minimize the conclusion instability problem across learners.MethodWe apply a Bellwether method comprising of three operators, namely, SORT+CLUSTER, GENERATE_TPM, and APPLY to sample the BMW from a pool of chronological projects from the Maxwell and International Software Benchmarking Standards Group (ISBSG) datasets. The sampled BMW is benchmarked against the entire collection of preprocessed projects, namely, growing portfolio to evaluate prediction and classification accuracy across a set of learners–ElasticNet regression, deep neural networks, and automatically transformed linear model.Results(1) BMW exists in the studied projects and (2) training the learners with a BMW of average window size 28.5%–75.5% of the growing portfolio (not older than 3 years) relatively minimizes the conclusion instability of prediction results.ConclusionWhen BMWs are available, we recommend their use for estimating the effort for a new project to minimize the conclusion instability problem.

Genome-Wide Association and Genomic Prediction of Growth Traits in the European Flat Oyster (Ostrea edulis).

The European flat oyster (Ostrea edulis) is a bivalve mollusc that was once widely distributed across Europe and represented an important food resource for humans for centuries. Populations of O. edulis experienced a severe decline across their biogeographic range mainly due to overexploitation and disease outbreaks. To restore the economic and ecological benefits of European flat oyster populations, extensive protection and restoration efforts are in place within Europe. In line with the increasing interest in supporting restoration and oyster farming through the breeding of stocks with enhanced performance, the present study aimed to evaluate the potential of genomic selection for improving growth traits in a European flat oyster population obtained from successive mass-spawning events. Four growth-related traits were evaluated: total weight (TW), shell height (SH), shell width (SW) and shell length (SL). The heritability of the growth traits was in the low-moderate range, with estimates of 0.45, 0.37, 0.22, and 0.32 for TW, SH, SW and SL, respectively. A genome-wide association analysis revealed a largely polygenic architecture for the four growth traits, with two distinct QTLs detected on chromosome 4. To investigate whether genomic selection can be implemented in flat oyster breeding at a reduced cost, the utility of low-density SNP panels was assessed. Genomic prediction accuracies using the full density panel were high (> 0.83 for all traits). The evaluation of the effect of reducing the number of markers used to predict genomic breeding values revealed that similar selection accuracies could be achieved for all traits with 2K SNPs as for a full panel containing 4,577 SNPs. Only slight reductions in accuracies were observed at the lowest SNP density tested (i.e., 100 SNPs), likely due to a high relatedness between individuals being included in the training and validation sets during cross-validation. Overall, our results suggest that the genetic improvement of growth traits in oysters is feasible. Nevertheless, and although low-density SNP panels appear as a promising strategy for applying GS at a reduced cost, additional populations with different degrees of genetic relatedness should be assessed to derive estimates of prediction accuracies to be expected in practical breeding programmes.