An Efficient Greedy Search Algorithm for High-dimensional Linear Discriminant Analysis.

High-dimensional classification is an important statistical problem that has applications in many areas. One widely used classifier is the Linear Discriminant Analysis (LDA). In recent years, many regularized LDA classifiers have been proposed to solve the problem of high-dimensional classification. However, these methods rely on inverting a large matrix or solving large-scale optimization problems to render classification rules-methods that are computationally prohibitive when the dimension is ultra-high. With the emergence of big data, it is increasingly important to develop more efficient algorithms to solve the high-dimensional LDA problem. In this paper, we propose an efficient greedy search algorithm that depends solely on closed-form formulae to learn a high-dimensional LDA rule. We establish theoretical guarantee of its statistical properties in terms of variable selection and error rate consistency; in addition, we provide an explicit interpretation of the extra information brought by an additional feature in a LDA problem under some mild distributional assumptions. We demonstrate that this new algorithm drastically improves computational speed compared with other high-dimensional LDA methods, while maintaining comparable or even better classification performance.

PENALIZED REGRESSION FOR MULTIPLE TYPES OF MANY FEATURES WITH MISSING DATA.

Recent technological advances have made it possible to measure multiple types of many features in biomedical studies. However, some data types or features may not be measured for all study subjects because of cost or other constraints. We use a latent variable model to characterize the relationships across and within data types and to infer missing values from observed data. We develop a penalized-likelihood approach for variable selection and parameter estimation and devise an efficient expectation-maximization algorithm to implement our approach. We establish the asymptotic properties of the proposed estimators when the number of features increases at a polynomial rate of the sample size. Finally, we demonstrate the usefulness of the proposed methods using extensive simulation studies and provide an application to a motivating multi-platform genomics study.

SEMIPARAMETRIC LATENT-CLASS MODELS FOR MULTIVARIATE LONGITUDINAL AND SURVIVAL DATA.

In long-term follow-up studies, data are often collected on repeated measures of multivariate response variables as well as on time to the occurrence of a certain event. To jointly analyze such longitudinal data and survival time, we propose a general class of semiparametric latent-class models that accommodates a heterogeneous study population with flexible dependence structures between the longitudinal and survival outcomes. We combine nonparametric maximum likelihood estimation with sieve estimation and devise an efficient EM algorithm to implement the proposed approach. We establish the asymptotic properties of the proposed estimators through novel use of modern empirical process theory, sieve estimation theory, and semiparametric efficiency theory. Finally, we demonstrate the advantages of the proposed methods through extensive simulation studies and provide an application to the Atherosclerosis Risk in Communities study.

Semiparametric single-index models for optimal treatment regimens with censored outcomes.

There is a growing interest in precision medicine, where a potentially censored survival time is often the most important outcome of interest. To discover optimal treatment regimens for such an outcome, we propose a semiparametric proportional hazards model by incorporating the interaction between treatment and a single index of covariates through an unknown monotone link function. This model is flexible enough to allow non-linear treatment-covariate interactions and yet provides a clinically interpretable linear rule for treatment decision. We propose a sieve maximum likelihood estimation approach, under which the baseline hazard function is estimated nonparametrically and the unknown link function is estimated via monotone quadratic B-splines. We show that the resulting estimators are consistent and asymptotically normal with a covariance matrix that attains the semiparametric efficiency bound. The optimal treatment rule follows naturally as a linear combination of the maximum likelihood estimators of the model parameters. Through extensive simulation studies and an application to an AIDS clinical trial, we demonstrate that the treatment rule derived from the single-index model outperforms the treatment rule under the standard Cox proportional hazards model.

Design and analysis of bridging studies with prior probabilities on the null and alternative hypotheses.

The pharmaceutical industry and regulatory agencies are increasingly interested in conducting bridging studies in order to bring an approved drug product from the original region (eg, United States or European Union) to a new region (eg, Asian-Pacific countries). In this article, we provide a new methodology for the design and analysis of bridging studies by assuming prior knowledge on how the null and alternative hypotheses in the original, foreign study are related to the null and alternative hypotheses in the bridging study and setting the type I error for the bridging study according to the strength of the foreign-study evidence. The new methodology accounts for randomness in the foreign-study evidence and controls the average type I error of the bridging study over all possibilities of the foreign-study evidence. In addition, the new methodology increases statistical power, when compared to approaches that do not use foreign-study evidence, and it allows for the possibility of not conducting the bridging study when the foreign-study evidence is unfavorable. Finally, we conducted extensive simulation studies to demonstrate the usefulness of the proposed methodology.

Silica-based colloid centrifugation enhances sperm quality in cockerel semen.

1. PercollTM is one of the most widely used colloid for animal sperm preparation. The aim of this study was to evaluate whether PercollTM colloid centrifugation could be practical to improve cockerel sperm quality, and to compare the effects of PercollTM single layer centrifugation (SLC) and density gradient centrifugation (DGC) in order to obtain the most optimal protocol for cockerel semen.2. In the experiment with PercollTM SLC for fresh semen, an increase of motile sperm was seen after PercollTM 80% SLC and 90% SLC was conducted, at levels of 28.8% and 30.2% respectively (P < 0.01). The increase of progressively motile sperm after PercollTM 80% SLC and 90% SLC was 177.2% and 202.4% respectively (P < 0.01). Meanwhile, for semen stored at 4°C for 24 h, the increase of motile sperm after PercollTM 70% SLC and 80% SLC was 41.2% and 44.0% (P < 0.01), and the increase of progressive sperm after PercollTM 70% SLC and 80% SLC was 71.3% and 83.1% respectively (P < 0.01). Both the percentage of motile sperm and progressive sperm of the fresh and stored cockerel semen after appropriate PercollTM SLC was significantly enhanced.3. Sperm membrane integrity did not show any decrease after PercollTM centrifugation compared with non-centrifuged semen, which suggested that the PercollTM centrifugation treatment in this study did not cause damage to cockerel sperm membranes.4. In the experiment regarding the comparison of PercollTM SLC and DGC with fresh semen, the increase of motile sperm after PercollTM 80% SLC, 90% SLC and 40%/80% DGC was 29.5%, 36.4%, and 25.0% respectively; and the increase of progressive sperm was 44.7%, 58.5%, and 54.7%, respectively. For semen stored at 4°C for 24 h, the increase of motile sperm after PercollTM 70% SLC, 80% SLC and 35%/70% DGC were 41.2%, 44.0%, and 26.4%; and the increase of progressive sperm was 71.3%, 83.1%, and 43.7%, respectively. There were no significant differences between the increase of sperm motility after PercollTM 80%, 90% SLC or PercollTM 40%/80% DGC in fresh cockerel semen. There was no significant difference between PercollTM 70%, 80% SLC and PercollTM 35%/70% in stored cockerel semen. There was a tendency for sperm recovery rates with PercollTM SLC to be higher than PercollTM DGC, although this did not reach statistical significance in this study.5. It was concluded that PercollTM SLC was more suitable for cockerel sperm separation than PercollTM DGC. The results suggested that PercollTM 80% SLC was the most optimal procedure to separate fresh cockerel sperm and PercollTM 70% SLC was the most optimal procedure to separate stored cockerel sperm. PercollTM SLC is more simple, user-friendly and economical and less time-consuming than DGC for cockerel semen processing.

Evaluation of cockerel spermatozoa viability and motility by a novel enzyme based cell viability assay.

1. The results of spermatozoa assessment by the WST-8 (2-[2-methoxy-4-nitrophenyl]-3-[4-nitrophenyl]-5-[2,4-disulfophenyl]-2H-tetrazolium, monosodium salt) assay, flow cytometry (FC) or computer-assisted sperm analysis (CASA) were compared. 2. Different live/killed ratios of cockerel semen were serially diluted to 120, 60, and 30 × 106 cells/ml, and each sample was analysed by (1) WST-8 assay at 0, 10, 20, 30, 40, 50, 60 min, (2) viability with FC, and (3) motility with CASA. 3. The WST-8 reduction rate was closely correlated with spermatozoa viability and motility. The optimal semen concentration for the WST-8 assay was 120 × 106 cells/ml, and the standard curves for spermatozoa viability and motility predictions, respectively, were yviability60 = 162.8x + 104.96 (R2 = 0.9594) after 60 min of incubation and ymotility40 = 225.09x + 96.299 (R2 = 0.8475) after 40 min of incubation. 4. It was concluded that the WST-8 assay is useful for the practical evaluation of cockerel spermatozoa viability and motility. Compared to FC and CASA, the WST-8 assay does not require expensive and complex instrumentation in the lab. Furthermore, one well of the WST-8 reaction can be used to predict spermatozoa viability and motility at the same time, which all lead it to be efficient and economical for semen quality assessment.

[Validation and optimization of the indicator system of risk assessment for mechanical cuts].

Objective: To validation and optimization the indicator system of risk assessment for mechanical cuts. Methods: The risk assessment index system of mechanical cutting injury established earlier was used to assess the risk of mechanical cutting injury in 40 cases of mechanical cutting injury registered from January 2015 to December 2017 and 40 similar positions without accidents in the same period. The multiple stepwise regression analysis was used to screen the indicator system, and to adjust the weight coefficient of each index. The total coincidence rate and Kappa value were compared between before and after optimization respectively. Results: The new index system has 3 first-class indicators, 10 second-class indicators and 14 three-class indicators, fewer than the old index system which has 3 first-class indicators, 10 second-class indicators, 34 three-class indicators. There three indicators have revamped in the first-class. The total of coincidence rates of the new and old indicator systems were 67.50% and 90.00%, the difference was statistically significant (P<0.01). The Kappa value were 0.35 and 0.80, respectively. Conclusion: The evaluation results with new indicator systems is more consistent with the actual hazard detection the the old indicator systems, and scientific, reasonable and practical, and the indicator system of risk assessment for mechanical cuts can be used for the risk assessment of mechanical cutting injuries.

Trans-ethnic analysis of metabochip data identifies two new loci associated with BMI.

OBJECTIVE: Body mass index (BMI) is commonly used to assess obesity, which is associated with numerous diseases and negative health outcomes. BMI has been shown to be a heritable, polygenic trait, with close to 100 loci previously identified and replicated in multiple populations. We aim to replicate known BMI loci and identify novel associations in a trans-ethnic study population. SUBJECTS: Using eligible participants from the Population Architecture using Genomics and Epidemiology consortium, we conducted a trans-ethnic meta-analysis of 102 514 African Americans, Hispanics, Asian/Native Hawaiian, Native Americans and European Americans. Participants were genotyped on over 200 000 SNPs on the Illumina Metabochip custom array, or imputed into the 1000 Genomes Project (Phase I). Linear regression of the natural log of BMI, adjusting for age, sex, study site (if applicable), and ancestry principal components, was conducted for each race/ethnicity within each study cohort. Race/ethnicity-specific, and combined meta-analyses used fixed-effects models. RESULTS: We replicated 15 of 21 BMI loci included on the Metabochip, and identified two novel BMI loci at 1q41 (rs2820436) and 2q31.1 (rs10930502) at the Metabochip-wide significance threshold (P<2.5 × 10-7). Bioinformatic functional investigation of SNPs at these loci suggests a possible impact on pathways that regulate metabolism and adipose tissue. CONCLUSION: Conducting studies in genetically diverse populations continues to be a valuable strategy for replicating known loci and uncovering novel BMI associations.

Improving the power to establish clinical similarity in a Phase 3 efficacy trial by incorporating prior evidence of analytical and pharmacokinetic similarity.

To improve patients' access to safe and effective biological medicines, abbreviated licensure pathways for biosimilar and interchangeable biological products have been established in the US, Europe, and other countries around the world. The US Food and Drug Administration and European Medicines Agency have published various guidance documents on the development and approval of biosimilars, which recommend a "totality-of-the-evidence" approach with a stepwise process to demonstrate biosimilarity. The approach relies on comprehensive comparability studies ranging from analytical and nonclinical studies to clinical pharmacokinetic/pharmacodynamic (PK/PD) and efficacy studies. A clinical efficacy study may be necessary to address residual uncertainty about the biosimilarity of the proposed product to the reference product and support a demonstration that there are no clinically meaningful differences. In this article, we propose a statistical strategy that takes into account the similarity evidence from analytical assessments and PK studies in the design and analysis of the clinical efficacy study in order to address residual uncertainty and enhance statistical power and precision. We assume that if the proposed biosimilar product and the reference product are shown to be highly similar with respect to the analytical and PK parameters, then they should also be similar with respect to the efficacy parameters. We show that the proposed methods provide correct control of the type I error and improve the power and precision of the efficacy study upon the standard analysis that disregards the prior evidence. We confirm and illustrate the theoretical results through simulation studies based on the biosimilars development experience of many different products.

Semiparametric regression for the weighted composite endpoint of recurrent and terminal events.

Recurrent event data are commonly encountered in clinical and epidemiological studies. A major complication arises when recurrent events are terminated by death. To assess the overall effects of covariates on the two types of events, we define a weighted composite endpoint as the cumulative number of recurrent and terminal events properly weighted by the relative severity of each event. We propose a semiparametric proportional rates model which specifies that the (possibly time-varying) covariates have multiplicative effects on the rate function of the weighted composite endpoint while leaving the form of the rate function and the dependence among recurrent and terminal events completely unspecified. We construct appropriate estimators for the regression parameters and the cumulative frequency function. We show that the estimators are consistent and asymptotically normal with variances that can be consistently estimated. We also develop graphical and numerical procedures for checking the adequacy of the model. We then demonstrate the usefulness of the proposed methods in simulation studies. Finally, we provide an application to a major cardiovascular clinical trial.

Sparse meta-analysis with high-dimensional data.

Meta-analysis plays an important role in summarizing and synthesizing scientific evidence derived from multiple studies. With high-dimensional data, the incorporation of variable selection into meta-analysis improves model interpretation and prediction. Existing variable selection methods require direct access to raw data, which may not be available in practical situations. We propose a new approach, sparse meta-analysis (SMA), in which variable selection for meta-analysis is based solely on summary statistics and the effect sizes of each covariate are allowed to vary among studies. We show that the SMA enjoys the oracle property if the estimated covariance matrix of the parameter estimators from each study is available. We also show that our approach achieves selection consistency and estimation consistency even when summary statistics include only the variance estimators or no variance/covariance information at all. Simulation studies and applications to high-throughput genomics studies demonstrate the usefulness of our approach.

Effect of prior deformation on microstructural development and Laves phase precipitation in high-chromium stainless steel.

This study investigated the influence of deformation on precipitation behaviour and microstructure change during annealing. Here, the prior deformation of high-chromium stainless steel was tensile deformation of 3%, 6% and 10%, and the specimens were then annealed at 700˚C for 10 h. The specimens were subsequently analyzed using backscattered electron image and electron backscattering diffraction measurements with SEM. Compared with the deformation microstructure, the grains revealed no preferred orientation. The precipitates of TiN and NbC were formed homogenously in the grain interior and at grain boundaries after annealing. Fine Laves phase precipitates were observed in grains and along subgrain boundaries as the deformation increased. Furthermore, the volume fraction of Laves phase increased, but the average particle diameter of precipitate was reduced as the deformation increased.

[The value of (18)F-FDG PET-CT imaging in predicting the malignant potential of GIST].

Objective: To evaluate the value of (18)F-FDG PET-CT in predicting the malignant potential of Gastrointestinal Stromal Tumors (GIST). Methods: The clinical and pathological features of 31 patients with GIST confirmed by surgery or biopsy were retrospectively analyzed. The malignant potential of GIST before treatment was assessed by (18)F-FDG PET-CT. The GIST risk classification was graded according to the Standard revised by the National Institutes of Health (NIH) in 2008. The relationship between the maximal standard uptake value (SUVmax) and GIST risk classification, tumor diameter, Ki-67 index, and mitotic count were analyzed respectively. The cut-off level of SUVmax for the diagnosis of malignant GIST was calculated from the Receiver Operating Characteristic (ROC) curve. Results: Among the 31 cases of GIST patients, 14 cases were gastric primary (stomach group) and 17 cases were nongastric primary (outside stomach group). The SUVmax, tumor diameter, Ki-67 index and mitotic count of the 31 patients were 8.21±4.68, (7.82±5.12)cm, (10.03±11.07)% and (12.29±10.55)/50 HPF, respectively. SUVmax was significantly correlated with GIST risk classification (r=0.727, P<0.01), but not with tumor diameter, Ki-67 index and mitotic count (r=0.348, r=0.284, r=0.290, P=0.055, P=0.121, P=0.114). The SUVmax, tumor diameter, Ki-67 index and mitotic count in the stomach group were 4.36±2.36, (6.08±4.31)cm, (3.43±3.03)% and (5.71±2.20)/50 HPF, respectively. SUVmax was significantly correlated with tumor diameter, GIST risk classification and Ki-67 index (r=0.682, r=0.868, r=0.732, P<0.01) but not with mitotic count (r=0.510, P=0.063). The SUVmax of the GIST in the gastric group and the outside gastric group were 4.36±2.36 and 10.68±5.50, respectively. The difference was statistically significant (P=0.001). The SUVmax in the malignant group of GIST (middle or high risk grade) was 8.90±4.89, which was significantly higher than 2.22±0.86 in the benign group (low or very low risk grade). The difference was statistically significant between the two group (P<0.01). ROC curve analysis showed that a SUVmax cut-off of 3.75 was the most sensitive for predicting malignant GIST. When the area under the curve of 0.969, the sensitivity was 84.6% and the specificity was 100%. Conclusions: The SUVmax was strongly correlated with the GIST risk category and also with the tumor diameter and Ki-67 index in the gastric primary GIST, so it can be used as an effective indicator in predicting malignant potential of GIST before treatment.

Polymorphic microsatellite loci isolated from Cervus unicolor (Cervidae) show inbreeding in a domesticated population of Taiwan Sambar deer.

Primers for eight microsatellites were developed; they successfully amplified DNA from 20 domesticated Formosan Sambar deer (Cervus unicolor swinhoei). All loci were polymorphic, with 10-19 alleles per locus. The average observed heterozygosity across loci and samples was 0.310, ranging from 0 to 0.750 at each locus. All loci but one, CU18, deviated from Hardy-Weinberg equilibrium due to excessive homozygosity in these domesticated broodstocks, reflecting inbreeding. These microsatellite loci will be useful, not only for assessment of population structure and genetic variability, but also for conservation of wild deer populations in Taiwan.

Survival analysis with incomplete genetic data.

Genetic data are now collected frequently in clinical studies and epidemiological cohort studies. For a large study, it may be prohibitively expensive to genotype all study subjects, especially with the next-generation sequencing technology. Two-phase sampling, such as case-cohort and nested case-control sampling, is cost-effective in such settings but entails considerable analysis challenges, especially if efficient estimators are desired. Another type of missing data arises when the investigators are interested in the haplotypes or the genetic markers that are not on the genotyping platform used for the current study. Valid and efficient analysis of such missing data is also interesting and challenging. This article provides an overview of these issues and outlines some directions for future research.

Multi-ancestry Genome-Wide Association Study of Early Childhood Caries.

Early childhood caries (ECC) is the most common non-communicable childhood disease. It is an important health problem with known environmental and social/behavioral influences that lacks evidence for specific associated genetic risk loci. To address this knowledge gap, we conducted a genome-wide association study of ECC in a multi-ancestry population of U.S. preschool-age children (n=6,103) participating in a community-based epidemiologic study of early childhood oral health. Calibrated examiners used ICDAS criteria to measure ECC with the primary trait using the dmfs index with decay classified as macroscopic enamel loss (ICDAS ≥3). We estimated heritability, concordance rates, and conducted genome-wide association analyses to estimate overall genetic effects; the effects stratified by sex, household water fluoride, and dietary sugar; and leveraged the combined gene/gene-environment effects using the 2-degree-of-freedom (2df) joint test. The common genetic variants explained 24% of the phenotypic variance (heritability) of the primary ECC trait and the concordance rate was higher with a higher degree of relatedness. We identified 21 novel non-overlapping genome-wide significant loci for ECC. Two loci, namely RP11-856F16 . 2 (rs74606067) and SLC41A3 (rs71327750) showed evidence of association with dental caries in external cohorts, namely the GLIDE consortium adult cohort (n=∼487,000) and the GLIDE pediatric cohort (n=19,000), respectively. The gene-based tests identified TAAR6 as a genome-wide significant gene. Implicated genes have relevant biological functions including roles in tooth development and taste. These novel associations expand the genomics knowledge base for this common childhood disease and underscore the importance of accounting for sex and pertinent environmental exposures in genetic investigations of oral health.

Checking semiparametric transformation models with censored data.

Semiparametric transformation models provide a very general framework for studying the effects of (possibly time-dependent) covariates on survival time and recurrent event times. Assessing the adequacy of these models is an important task because model misspecification affects the validity of inference and the accuracy of prediction. In this paper, we introduce appropriate time-dependent residuals for these models and consider the cumulative sums of the residuals. Under the assumed model, the cumulative sum processes converge weakly to zero-mean Gaussian processes whose distributions can be approximated through Monte Carlo simulation. These results enable one to assess, both graphically and numerically, how unusual the observed residual patterns are in reference to their null distributions. The residual patterns can also be used to determine the nature of model misspecification. Extensive simulation studies demonstrate that the proposed methods perform well in practical situations. Three medical studies are provided for illustrations.

Genetic diversity of Vietnamese domestic chicken populations as decision-making support for conservation strategies.

The aims of this study were to assess the genetic diversity of 17 populations of Vietnamese local chickens (VNN) and one Red Jungle Fowl population, together with six chicken populations of Chinese origin (CNO), and to provide priorities supporting the conservation of genetic resources using 20 microsatellites. Consequently, the VNN populations exhibited a higher diversity than did CNO populations in terms of number of alleles but showed a slightly lower observed heterozygosity. The VNN populations showed in total seven private alleles, whereas no CNO private alleles were found. The expected heterozygosity of 0.576 in the VNN populations was higher than the observed heterozygosity of 0.490, leading to heterozygote deficiency within populations. This issue could be partly explained by the Wahlund effect due to fragmentation of several populations between chicken flocks. Molecular analysis of variance showed that most of genetic variation was found within VNN populations. The Bayesian clustering analysis showed that VNN and CNO chickens were separated into two distinct groups with little evidence for gene flow between them. Among the 24 populations, 13 were successfully assigned to their own cluster, whereas the structuring was not clear for the remaining 11 chicken populations. The contributions of 24 populations to the total genetic diversity were mostly consistent across two approaches, taking into account the within- and between-populations genetic diversity and allelic richness. The black H'mong, Lien Minh, Luong Phuong and Red Jungle Fowl were ranked with the highest priorities for conservation according to Caballero and Toro's and Petit's approaches. In conclusion, a national strategy needs to be set up for Vietnamese chicken populations, with three main components: conservation of high-priority breeds, within-breed management with animal exchanges between flocks to avoid Wahlund effect and monitoring of inbreeding rate.