Simulations of multiple breeding strategy scenarios in common bean for assessing genomic selection accuracy and model updating.

The aim of this study was to evaluate the accuracy of the ridge regression best linear unbiased prediction model across different traits, parent population sizes, and breeding strategies when estimating breeding values in common bean (Phaseolus vulgaris). Genomic selection was implemented to make selections within a breeding cycle and compared across five different breeding strategies (single seed descent, mass selection, pedigree method, modified pedigree method, and bulk breeding) following 10 breeding cycles. The model was trained on a simulated population of recombinant inbreds genotyped for 1010 single nucleotide polymorphism markers including 38 known quantitative trait loci identified in the literature. These QTL included 11 for seed yield, eight for white mold disease incidence, and 19 for days to flowering. Simulation results revealed that realized accuracies fluctuate depending on the factors investigated: trait genetic architecture, breeding strategy, and the number of initial parents used to begin the first breeding cycle. Trait architecture and breeding strategy appeared to have a larger impact on accuracy than the initial number of parents. Generally, maximum accuracies (in terms of the correlation between true and estimated breeding value) were consistently achieved under a mass selection strategy, pedigree method, and single seed descent method depending on the simulation parameters being tested. This study also investigated model updating, which involves retraining the prediction model with a new set of genotypes and phenotypes that have a closer relation to the population being tested. While it has been repeatedly shown that model updating generally improves prediction accuracy, it benefited some breeding strategies more than others. For low heritability traits (e.g., yield), conventional phenotype-based selection methods showed consistent rates of genetic gain, but genetic gain under genomic selection reached a plateau after fewer cycles. This plateauing is likely a cause of faster fixation of alleles and a diminishing of genetic variance when selections are made based on estimated breeding value as opposed to phenotype.

Simulations of rate of genetic gain in dry bean breeding programs.

KEY MESSAGE: A reference study for breeders aiming at maximizing genetic gain in common bean. Depending on trait heritability and genetic architecture, conventional approaches may provide an advantage over other frameworks. Dry beans (Phaseolus vulgaris L.) are a nutrient dense legume that is consumed by developed and developing nations around the world. The progress to improve this crop has been quite steady. However, with the continued rise in global populations, there are demands to expedite genetic gains. Plant breeders have been at the forefront at increasing yields in the common bean. As breeding programs are both time-consuming and resource intensive, resource allocation must be carefully considered. To assist plant breeders, computer simulations can provide useful information that may then be applied to the real world. This study evaluated multiple breeding scenarios in the common bean and involved five selection strategies, three breeding frameworks, and four different parental population sizes. In addition, the breeding scenarios were implemented in three different traits: days to flowering, white mold tolerance, and seed yield. Results from the study reflect the complexity of breeding programs, with the optimal breeding scenario varying based on trait being selected. Relative genetic gains per cycle of up to 8.69% for seed yield could be obtained under the use of the optimal breeding scenario. Principal component analyses revealed similarity between strategies, where single seed descent and the modified pedigree method would often aggregate. As well, clusters in the direction of the Hamming distance eigenvector are a good indicator of poor performance in a strategy.

Deterministic and stochastic modelling of impacts from genomic selection and phenomics on genetic gain for perennial ryegrass dry matter yield.

Increasing the efficiency of current forage breeding programs through adoption of new technologies, such as genomic selection (GS) and phenomics (Ph), is challenging without proof of concept demonstrating cost effective genetic gain (∆G). This paper uses decision support software DeltaGen (tactical tool) and QU-GENE (strategic tool), to model and assess relative efficiency of five breeding methods. The effect on ∆G and cost ($) of integrating GS and Ph into an among half-sib (HS) family phenotypic selection breeding strategy was investigated. Deterministic and stochastic modelling were conducted using mock data sets of 200 and 1000 perennial ryegrass HS families using year-by-season-by-location dry matter (DM) yield data and in silico generated data, respectively. Results demonstrated short (deterministic)- and long-term (stochastic) impacts of breeding strategy and integration of key technologies, GS and Ph, on ∆G. These technologies offer substantial improvements in the rate of ∆G, and in some cases improved cost-efficiency. Applying 1% within HS family GS, predicted a 6.35 and 8.10% ∆G per cycle for DM yield from the 200 HS and 1000 HS, respectively. The application of GS in both among and within HS selection provided a significant boost to total annual ∆G, even at low GS accuracy rA of 0.12. Despite some reduction in ∆G, using Ph to assess seasonal DM yield clearly demonstrated its impact by reducing cost per percentage ∆G relative to standard DM cuts. Open-source software tools, DeltaGen and QuLinePlus/QU-GENE, offer ways to model the impact of breeding methodology and technology integration under a range of breeding scenarios.

Maternal and infant predictors of infant mortality in California, 2007-2015.

OBJECTIVE: To identify current maternal and infant predictors of infant mortality, including maternal sociodemographic and economic status, maternal perinatal smoking and obesity, mode of delivery, and infant birthweight and gestational age. METHODS: This retrospective study analyzed data from the linked birth and infant death files (birth cohort) and live births from the Birth Statistical Master files (BSMF) in California compiled by the California Department of Public Health for 2007-2015. The birth cohort study comprised 4,503,197 singleton births including 19,301 infant deaths during the nine-year study period. A subpopulation to study fetal growth consisted of 4,448,300 birth cohort records including 13,891 infant deaths. RESULTS: The infant mortality rate (IMR) for singleton births decreased linearly (p <0.001) from 4.68 in 2007 to 3.90 (per 1,000 live births) in 2015. However, significant disparities in IMR were uncovered in different population groups depending upon maternal sociodemographic and economic characteristics and maternal characteristics during pregnancy. Children of African American women had almost twice the risk of infant mortality when compared with children of White women (AOR 2.12; 95% CI, 1.98-2.27; p<0.001). Infants of women with Bachelor's degrees or higher were 89% less likely to die (AOR 1.89; 95% CI, 1.76-2.04; p<0.001) when compared to infants of women with education less than high school. Infants of maternal smokers were 75% more likely to die (AOR 1.75; 95% CI, 1.58-1.93; p<0.001) than infants of nonsmokers. Infants of women who were overweight and obese during pregnancy accounted for 55% of IMR over all women in the study. More than half of the infant deaths were to children of women with lower socioeconomic status; infants of WIC participants were 59% more likely to die (AOR 1.59; 95% CI, 1.52-1.67; p<0.001) than infants of non-WIC participants. With respect to infant predictors, infants born with LBW or PTB were more than six times (AOR 6.29; 95% CI, 5.90-6.70; p<0.001) and almost four times (AOR 3.95; 95% CI, 3.73-4.19; p<0.001) more likely to die than infants who had normal births, respectively. SGA and LGA infants were more than two times (AOR 2.03; 95% CI, 1.92-2.15; p<0.001) and 41% (AOR 1.41; 95% CI, 1.32-1.52; p<0.001) more likely to die than AGA infants, respectively. CONCLUSIONS: While the overall IMR in California is declining, wide disparities in death rates persist in different groups, and these disparities are increasing. Our data indicate that maternal sociodemographic and economic factors, as well as maternal prepregnancy obesity and smoking during pregnancy, have a prominent effect on IMR though no causality can be inferred with the current data. These predictors are not typically addressed by direct medical care. Infant factors with a major effect on IMR are birthweight and gestational age-predictors that are addressed by active medical services. The highest value interventions to reduce IMR may be social and public health initiatives that mitigate disparities in sociodemographic, economic and behavioral risks for mothers.

Smoking during Pregnancy and Adverse Birth and Maternal Outcomes in California, 2007 to 2016.

OBJECTIVE: This study aimed to determine associations between maternal cigarette smoking and adverse birth and maternal outcomes. STUDY DESIGN: This is a 10-year population-based retrospective cohort study including 4,971,896 resident births in California. Pregnancy outcomes of maternal smokers were compared with those of nonsmokers. The outcomes of women who stopped smoking before or during various stages of pregnancy were also investigated. RESULTS: Infants of women who smoked during pregnancy were twice as likely to have low birth weight (LBW) and be small for gestational age (SGA), 57% more likely to have very LBW (VLBW) or be a preterm birth (PTB), and 59% more likely to have a very PTB compared with infants of nonsmokers. During the study period, a significant widening of gaps developed in both rates of LBW and PTB and the percentage of SGA between infants of maternal smokers and nonsmokers. CONCLUSION: Smoking during pregnancy is associated with a significantly increased risk of adverse birth and maternal outcomes, and differences in rates of LBW, PTB, and SGA between infants of maternal smokers and nonsmokers increased during this period. Stopping smoking before pregnancy or even during the first trimester significantly decreased the infant risks of LBW, PTB, SGA, and the maternal risk for cesarean delivery.

Trends in maternal prepregnancy body mass index (BMI) and its association with birth and maternal outcomes in California, 2007-2016: A retrospective cohort study.

OBJECTIVE: To determine recent trends in maternal prepregnancy body mass index (BMI) and to quantify its association with birth and maternal outcomes. METHODS: A population-based retrospective cohort study included resident women with singleton births in the California Birth Statistical Master Files (BSMF) database from 2007 to 2016. There were 4,621,082 women included out of 5,054,968 women registered in the database. 433,886 (8.6%) women were excluded due to invalid or missing information for BMI. Exposures were underweight (BMI < 18.5 kg/m2), normal weight (18.5-24.9 kg/m2), overweight (25.0-29.9 kg/m2), and obese (≥ 30 kg/m2) at the onset of pregnancy. Obesity was subcategorized into class I (30.0-34.9 kg/m2), class II (35.0-39.9 kg/m2), and class III (≥ 40 kg/m2), while adverse outcomes examined were low birth weight (LBW), very low birth weight (VLBW), macrosomic births, preterm birth (PTB), very preterm birth (VPTB), small-for-gestational-age birth (SGA), large-for-gestational-age birth (LGA), and cesarean delivery (CD). Descriptive analysis, simple linear regression, and multivariate logistic regression were performed, and adjusted odds ratios (AORs) with 95% confidence intervals (CIs) for associations were estimated. RESULTS: Over the ten-year study period, the prevalence of underweight and normal weight women at time of birth declined by 10.6% and 9.7%, respectively, while the prevalence of overweight and obese increased by 4.3% and 22.9%, respectively. VLBW increased significantly with increasing BMI, by 24% in overweight women and by 76% in women with class III obesity from 2007 to 2016. Women with class III obesity also had a significant increase in macrosomic birth (170%) and were more likely to deliver PTB (33%), VPTB (66%), LGA (231%), and CD (208%) than women with a normal BMI. However, obese women were less likely to have SGA infants; underweight women were 51% more likely to have SGA infants than women with a normal BMI. CONCLUSIONS: In California from 2007 to 2016, there was a declining trend in women with prepregnancy normal weight, and a rising trend in overweight and obese women, particularly obesity class III. Both extremes of prepregnancy BMI were associated with an increased incidence of adverse neonatal outcomes; however, the worse outcomes were prominent in those women classified as obese.

QuLinePlus: extending plant breeding strategy and genetic model simulation to cross-pollinated populations-case studies in forage breeding.

Plant breeders are supported by a range of tools that assist them to make decisions about the conduct or design of plant breeding programs. Simulations are a strategic tool that enables the breeder to integrate the multiple components of a breeding program into a number of proposed scenarios that are compared by a range of statistics measuring the efficiency of the proposed systems. A simulation study for the trait growth score compared two major strategies for breeding forage species, among half-sib family selection and among and within half-sib family selection. These scenarios highlighted new features of the QuLine program, now called QuLinePlus, incorporated to enable the software platform to be used to simulate breeding programs for cross-pollinated species. Each strategy was compared across three levels of half-sib family mean heritability (0.1, 0.5, and 0.9), across three sizes of the initial parental population (10, 50, and 100), and across three genetic effects models (fully additive model, a mixture of additive, partial and over dominance model, and a mixture of partial dominance and over dominance model). Among and within half-sib selection performed better than among half-sib selection for all scenarios. The new tools introduced into QuLinePlus should serve to accurately compare among methods and provide direction on how to achieve specific goals in the improvement of plant breeding programs for cross breeding species.

Temporal trends, patterns, and predictors of preterm birth in California from 2007 to 2016, based on the obstetric estimate of gestational age.

BACKGROUND: Preterm birth (PTB) is associated with increased infant mortality, and neurodevelopmental abnormalities among survivors. The aim of this study is to investigate temporal trends, patterns, and predictors of PTB in California from 2007 to 2016, based on the obstetric estimate of gestational age (OA). METHODS: A retrospective cohort study evaluated 435,280 PTBs from the 5,137,376 resident live births (8.5%) documented in the California Birth Statistical Master Files (BSMF) from 2007 to 2016. The outcome variable was PTB; the explanatory variables were birth year, maternal characteristics and health behaviors. Descriptive statistics and logistic regression analysis were used to identify subgroups with significant risk factors associated with PTB. Small for gestational age (SGA), appropriate for gestational age (AGA) and large for gestational age (LGA) infants were identified employing gestational age based on obstetric estimates and further classified by term and preterm births, resulting in six categories of intrauterine growth. RESULTS: The prevalence of PTB in California decreased from 9.0% in 2007 to 8.2% in 2014, but increased during the last 2 years, 8.4% in 2015 and 8.5% in 2016. Maternal age, education level, race and ethnicity, smoking during pregnancy, and parity were significant risk factors associated with PTB. The adjusted odds ratio (AOR) showed that women in the oldest age group (40-54 years) were almost twice as likely to experience PTB as women in the 20- to 24-year reference age group. The prevalence of PTB was 64% higher in African American women than in Caucasian women. Hispanic women showed less disparity in the prevalence of PTB based on education and socioeconomic level. The analysis of interactions between maternal characteristics and perinatal health behaviors showed that Asian women have the highest prevalence of PTB in the youngest age group (< 20 years; AOR, 1.40; 95% confidence interval (CI), 1.28-1.54). Pacific Islander, American Indian, and African American women ≥40 years of age had a greater than two-fold increase in the prevalence of PTB compared with women in the 20-24 year age group. Compared to women in the Northern and Sierra regions, women in the San Joaquin Valley were 18%, and women in the Inland Empire and San Diego regions 13% more likely to have a PTB. Women who smoked during both the first and second trimesters were 57% more likely to have a PTB than women who did not smoke. Compared to women of normal prepregnancy weight, underweight women and women in obese class III were 23 and 33% more likely to experience PTB respectively. CONCLUSIONS: Implementation of public health initiatives focusing on reducing the prevalence of PTB should focus on women of advanced maternal age and address race, ethnic, and geographic disparities. The significance of modifiable maternal perinatal health behaviors that contribute to PTB, e.g. smoking during pregnancy and prepregnancy obesity, need to be emphasized during prenatal care.

Recent trends, risk factors, and disparities in low birth weight in California, 2005-2014: a retrospective study.

BACKGROUND: Low birth weight (LBW) is a leading risk factor for infant morbidity and mortality in the United States. There are large disparities in the prevalence of LBW by race and ethnicity, especially between African American and White women. Despite extensive research, the practice of clinical and public health, and policies devoted to reducing the number of LBW infants, the prevalence of LBW has remained unacceptably and consistently high. There have been few detailed studies identifying the factors associated with LBW in California, which is home to a highly diverse population. The aim of this study is to investigate recent trends in the prevalence of LBW infants (measured as a percentage) and to identify risk factors and disparities associated with LBW in California. METHODS: A retrospective cohort study included data on 5,267,519 births recorded in the California Birth Statistical Master Files for the period 2005-2014. These data included maternal characteristics, health behaviors, information on health insurance, prenatal care use, and parity. Logistic regression models identified significant risk factors associated with LBW. Using gestational age based on obstetric estimates (OA), small for gestational age (SGA), appropriate for gestational age (AGA) and large for gestational age (LGA) infants were identified for the periods 2007-2014. RESULTS: The number of LBW infants declined, from 37,603 in 2005 to 33,447 in 2014. However, the prevalence of LBW did not change significantly (6.9% in 2005 to 6.7% in 2014). The mean maternal age at first delivery increased from 25.7 years in 2005 to 27.2 years in 2014. The adjusted odds ratio showed that women aged 40 to 54 years were twice as likely to have an LBW infant as women in the 20 to 24 age group. African American women had a persistent 2.4-fold greater prevalence of having an LBW infant compared with white women. Maternal age was a significant risk factor for LBW regardless of maternal race and ethnicity or education level. During the period 2017-2014, 5.4% of the singleton births at 23-41 weeks based on OE of gestational age were SGA infants (preterm SGA + term SGA). While all the preterm SGA infants were LBW, both preterm AGA and term SGA infants had a higher prevalence of LBW. CONCLUSIONS: In California, during the 10 years from 2005 to 2014, there was no significant decline in the prevalence of LBW. However, maternal age was a significant risk factor for LBW regardless of maternal race and ethnicity or education level. Therefore, there may be opportunities to reduce the prevalence of LBW by reducing disparities and improving birth outcomes for women of advanced maternal age.

A Genomic Selection Index Applied to Simulated and Real Data.

A genomic selection index (GSI) is a linear combination of genomic estimated breeding values that uses genomic markers to predict the net genetic merit and select parents from a nonphenotyped testing population. Some authors have proposed a GSI; however, they have not used simulated or real data to validate the GSI theory and have not explained how to estimate the GSI selection response and the GSI expected genetic gain per selection cycle for the unobserved traits after the first selection cycle to obtain information about the genetic gains in each subsequent selection cycle. In this paper, we develop the theory of a GSI and apply it to two simulated and four real data sets with four traits. Also, we numerically compare its efficiency with that of the phenotypic selection index (PSI) by using the ratio of the GSI response over the PSI response, and the PSI and GSI expected genetic gain per selection cycle for observed and unobserved traits, respectively. In addition, we used the Technow inequality to compare GSI vs. PSI efficiency. Results from the simulated data were confirmed by the real data, indicating that GSI was more efficient than PSI per unit of time.

Prediction of genetic values of quantitative traits in plant breeding using pedigree and molecular markers.

The availability of dense molecular markers has made possible the use of genomic selection (GS) for plant breeding. However, the evaluation of models for GS in real plant populations is very limited. This article evaluates the performance of parametric and semiparametric models for GS using wheat (Triticum aestivum L.) and maize (Zea mays) data in which different traits were measured in several environmental conditions. The findings, based on extensive cross-validations, indicate that models including marker information had higher predictive ability than pedigree-based models. In the wheat data set, and relative to a pedigree model, gains in predictive ability due to inclusion of markers ranged from 7.7 to 35.7%. Correlation between observed and predictive values in the maize data set achieved values up to 0.79. Estimates of marker effects were different across environmental conditions, indicating that genotype × environment interaction is an important component of genetic variability. These results indicate that GS in plant breeding can be an effective strategy for selecting among lines whose phenotypes have yet to be observed.