Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “model selection”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 1,261 records · Page 70Linked to original sources
PubMed2000

Comparing the agreement among alternative models in evaluating HMO efficiency.

OBJECTIVE: To describe the efficiency of HMOs and to test the robustness of these findings across alternative models of efficiency. This study examines whether these models, when constructed in parallel to use the same information, provide researchers with the same insights and identify the same trends. DATA SOURCES: A data set containing 585 HMOs operating from 1985 through 1994. Variables include enrollment, utilization, and financial information compiled primarily from Health Care Investment Analysts, InterStudy HMO Census, and Group Health Association of America. STUDY DESIGN: We compute three estimates of efficiency for each HMO and compare the results in terms of individual performance and industry-wide trends. The estimates are then regressed against measures of case mix, quality, and other factors that may be related to the model estimates. PRINCIPAL FINDINGS: The three models identify similar trends for the HMO industry as a whole; however, they assess the relative technical efficiency of individual firms differently. Thus, these techniques are limited for either benchmarking or setting rates because the firms identified as efficient may be a consequence of model selection rather than actual performance. CONCLUSIONS: The estimation technique to evaluate efficient firms can affect the findings themselves. The implications are relevant not only for HMOs, but for efficiency analyses in general. Concurrence among techniques is no guarantee of accuracy, but it is reassuring; conversely, radically distinct inferences across models can be a warning to temper research conclusions.

Efficiency, Organizational↗
PubMed2000

Summarizing the predictive power of a generalized linear model.

This paper studies summary measures of the predictive power of a generalized linear model, paying special attention to a generalization of the multiple correlation coefficient from ordinary linear regression. The population value is the correlation between the response and its conditional expectation given the predictors, and the sample value is the correlation between the observed response and the model predicted value. We compare four estimators of the measure in terms of bias, mean squared error and behaviour in the presence of overparameterization. The sample estimator and a jack-knife estimator usually behave adequately, but a cross-validation estimator has a large negative bias with large mean squared error. One can use bootstrap methods to construct confidence intervals for the population value of the correlation measure and to estimate the degree to which a model selection procedure may provide an overly optimistic measure of the actual predictive power.

Bias↗
PubMed2006

An ordered generalised extreme value model with application to alcohol consumption in Australia.

An Ordered Generalised Extreme Value (OGEV) model by Small (1987) is proposed for application to ordered discrete choice data. Relative to conventional Ordered Probit/Logit (OP/OL) and Multinomial Logit (MNL) models, the OGEV model is flexible, is defined by random utility maximization, and allows for correlation across choices via unobservable individual characteristics according to locations of the choices in the ordering. The OGEV model is applied to unit record data from Australia to study the impacts of prices, income and demographic characteristics on levels of alcohol consumption. Model selection analysis suggests that OGEV is preferred to both OP and MNL for the application.

Adolescent↗
PubMed1982

Neurochemical correlates of the kindling model of epilepsy.

Repeated electrical stimulation of the brain can produce many epileptogenic effects including those which characterize the kindling model. Kindling stimulation, by definition, changes the brain is such a way that formerly subconvulsive stimuli can elicit electrographic and convulsive seizure activity. In addition, the kindled animal becomes more susceptible to many, but not all, other types of seizures. These facts suggest that kindling produces brain changes which may selectively model some types of epileptiform excitability. In order to understand the basis for such changes numerous neurochemical studies have been attempted in the last few years. Although many changes have been demonstrated to be produced by kindling, few studies have been designed to specifically examine the long-lasting (permanent) neurochemical correlates of kindling stimulation. In this review, neurochemical data relevant to kindling are presented and discussed in terms of their possible significance to the seizure susceptibility changes produced by kindling.

Acetylcholine↗
PubMed2005

The fumarate sensor DcuS: progress in rapid protein fold elucidation by combining protein structure prediction methods with NMR spectroscopy.

We illustrate how moderate resolution protein structures can be rapidly obtained by interlinking computational prediction methodologies with un- or partially assigned NMR data. To facilitate the application of our recently described method of ranking and subsequent refining alternative structural models using unassigned NMR data [Proc. Natl. Acad. Sci. USA 100 (2003) 15404] for such "structural genomics"-type experiments it is combined with protein models from several prediction techniques, enhanced to utilize partial assignments, and applied on a protein with an unknown structure and fold. From the original NMR spectra obtained for the 140 residue fumarate sensor DcuS, 1100 1H, 13C, and 15N chemical shift signals, 3000 1H-1H NOESY cross peak intensities, and 209 backbone residual dipolar couplings were extracted and used to rank models produced by de novo structure prediction and comparative modeling methods. The ranking proceeds in two steps: first, an optimal assignment of the NMR peaks to atoms is found for each model independently, and second, the models are ranked based on the consistency between the NMR data and the model assuming these optimal assignments. The low-resolution model selected using this ranking procedure had the correct overall fold and a global backbone RMSD of 6.0 angstrom, and was subsequently refined to 3.7 angstrom RMSD. With the incorporation of a small number of NOE and residual dipolar coupling constraints available very early in the traditional spectral assignment process, a model with an RMSD of 2.8 angstrom could rapidly be built. The ability to generate moderate resolution models within days of NMR data collection should facilitate large scale NMR structure determination efforts.

Algorithms↗
PubMed2005

Towards a quantitative understanding of horizontal gene transfer: a kinetic model.

In this work, we present a simple kinetic model of horizontal gene transfer. It describes the processes of gene duplication, mutation, gene transfer and the regulation of the total size of the genome for genetically homogeneous prokaryotic species or strains. The emerging nonlinear system of first-order differential equations can be linearized at the stationary point. For selected models, we give an analytical solution for the number of foreign and native genes within a species. We identify a regime characterized by a fast gene transfer rate and species with a mixed genome, a slow gene transfer regime with pure organisms, and a crossover region. The data are compared to experiments, and the biological implications of our model are discussed.

Animals↗
PubMedDate not supplied

Using pharmacophore models to gain insight into structural binding and virtual screening: an application study with CDK2 and human DHFR.

This study provides results from two case studies involving the application of the HypoGenRefine algorithm within Catalyst for the automated generation of excluded volume from ligand information alone. A limitation of pharmacophore feature hypothesis alone is that activity prediction is based purely on the presence and arrangement of pharmacophoric features; steric effects remained unaccounted. Recently reported studies have illustrated the usefulness of combining excluded volumes to the pharmacophore models. In general, these excluded volumes attempt to penalize molecules occupying steric regions that are not occupied by active molecules. The HypoGenRefine algorithm in Catalyst accounts for steric effects on activity, based on the targeted addition of excluded volume features to the pharmacophores. The automated inclusion of excluded volumes to pharmacophore models has been applied to two systems: CDK2 and human DHFR. These studies are used as examples to illustrate how ligands could bind in the protein active site with respect to allowed and disallowed binding regions. Additionally, automated refinement of the pharmacophore with these excluded volume features provides a more selective model to reduce false positives and a better enrichment rate in virtual screening.

Algorithms↗
PubMed1999

Bayesian inference for prevalence in longitudinal two-phase studies.

We consider Bayesian inference and model selection for prevalence estimation using a longitudinal two-phase design in which subjects initially receive a low-cost screening test followed by an expensive diagnostic test conducted on several occasions. The change in the subject's diagnostic probability over time is described using four mixed-effects probit models in which the subject-specific effects are captured by latent variables. The computations are performed using Markov chain Monte Carlo methods. These models are then compared using the deviance information criterion. The methodology is illustrated with an analysis of alcohol and drug use in adolescents using data from the Great Smoky Mountains Study.

Adolescent↗
PubMed2005

Extension of multifactor dimensionality reduction for identifying multilocus effects in the GAW14 simulated data.

The multifactor dimensionality reduction (MDR) is a model-free approach that can identify gene x gene or gene x environment effects in a case-control study. Here we explore several modifications of the MDR method. We extended MDR to provide model selection without crossvalidation, and use a chi-square statistic as an alternative to prediction error (PE). We also modified the permutation test to provide different levels of stringency. The extended MDR (EMDR) includes three permutation tests (fixed, non-fixed, and omnibus) to obtain p-values of multilocus models. The goal of this study was to compare the different approaches implemented in the EMDR method and evaluate the ability to identify genetic effects in the Genetic Analysis Workshop 14 simulated data. We used three replicates from the simulated family data, generating matched pairs from family triads. The results showed: 1) chi-square and PE statistics give nearly consistent results; 2) results of EMDR without cross-validation matched that of EMDR with 10-fold cross-validation; 3) the fixed permutation test reports false-positive results in data from loci unrelated to the disease, but the non-fixed and omnibus permutation tests perform well in preventing false positives, with the omnibus test being the most conservative. We conclude that the non-cross-validation test can provide accurate results with the advantage of high efficiency compared to 10-cross-validation, and the non-fixed permutation test provides a good compromise between power and false-positive rate.

Computer Simulation↗
PubMed1999

Relationships between exposure to saquinavir monotherapy and antiviral response in HIV-positive patients.

OBJECTIVE: The aim of this study was to confirm the most appropriate dosage of a new soft gelatin capsule (SGC) formulation of the HIV protease inhibitor saquinavir by investigating the relationships between systemic (plasma) exposure to saquinavir and plasma HIV RNA and CD4+ cell counts using empirical mathematical modelling. DESIGN AND SETTING: A randomised, non-blind, multicentre, dose-ranging 8-week study of monotherapy with 400, 800 or 1200 mg of saquinavir-SGC or 600 mg of the hard gelatin capsule (HGC) formulation, both administered 3 times daily, was carried out in protease inhibitor-naive, HIV-positive adults. Two surrogate markers of response, plasma HIV RNA level and CD4+ cell count, were fitted to 2 measures of systemic drug exposure, the area under the plasma concentration-time curve (AUC) and trough plasma concentration (Cmin), using 6 exposure-response models of progressively increasing complexity. Akaike and Schwarz model selection criteria were applied to determine the most effective pharmacokinetic-pharmacodynamic relationship. RESULTS: A total of 88 patients were randomised; pharmacokinetic and pharmacodynamic data were available for 84 patients. In terms of plasma HIV RNA, pharmacokinetic-pharmacodynamic relationships were best described by a 2-parameter maximum effect (Emax) model, which predicted a typical maximum reduction in viral load of 1.94 log10 copies/ml [coefficient of variation (CV) 12%], with a half-maximal antiviral response occurring at a Cmin of 50 micrograms/L (CV 40%). Saquinavir-SGC 1200 mg administered 3 times daily produced a median AUC to 24 hours (AUC24) of approximately 20,000 micrograms/L.h, corresponding to 85% of the maximum achievable antiviral effect as defined by the model. None of the models yielded a satisfactory fit for CD4+ cell count. CONCLUSION: Empirical mathematical modelling confirmed that, when administered 3 times daily, the optimum dose of saquinavir-SGC is 1200 mg, corresponding to 3600 mg/day.

Adult↗
PubMed1990

Microcomputer-assisted multivariate survival data analysis using Cox's proportional hazards regression model.

We describe a microcomputer program (COXSURV) for proportional hazards multiple regression analysis of survival and other failure-time data generated in clinical trials and in retrospective clinical epidemiology studies. COXSURV is menu-driven and has powerful variable factoring and data exploratory capabilities for multivariate modeling. A batch mode allows automatic uni- or multivariate analyses for confounder summarization. Model selection for predictive purposes is possible through a step-up algorithm. The partial likelihood method used in the program allows the use of either discrete or continuous time scales by treating tied uncensored observations by either the exact method or by a robust approximation method. The program calculates most standard model fitting statistics for either overall or stratified analyses and uses data layout files compatible with those of other related epidemiologic analysis software.

Algorithms↗
PubMed2005

The effect of solvent polarity on the accumulation of leachables from pharmaceutical product containers.

Material/water equilibrium interaction constants (E(b)) were determined for 12 organic model solutes and a plastic material used in pharmaceutical product containers (non-PVC polyolefin). An excellent correlation was obtained between the measured interaction constants and the organic solute's octanol/water partition coefficient. The effect of solvent polarity on E(b) was assessed by examining the interaction between the plastic and selected model solutes in binary ethanol/water mixtures. In general, logE(b) could be linearily related to the polarity of the ethanol/water mixture. This information, coupled with the interaction model, was used to estimate the levels to which container leachables could accumulate in contacted solutions. Such estimates were made for six known leachables of the polyolefin material and compared to the leachable's measured accumulation levels in binary ethanol/water systems. In general, the accumulation level of the leachables increased with increasing solution polarity. For most of the leachables, the measured accumulation level was less than the calculated levels, suggesting that equilibrium was not achieved in the leaching portion of this study. This lack of equilibrium is attributable to the layered structure of the material studied, as such layering retards the migration of the leachables that are derived from the material's non-solution contact layers.

Chromatography, High Pressure Liquid↗
PubMed2005

Modeling of slow glutamate diffusion and AMPA receptor activation in the cerebellar glomerulus.

Synaptic conductances are influenced markedly by the geometry of the space surrounding the synapse since the transient glutamate concentration in the synaptic cleft is determined by this geometry. Our paper is an attempt to understand the reasons for slow glutamate diffusion in the cerebellar glomerulus, a structure situated around the enlarged mossy fiber terminal in the cerebellum and surrounded by a glial sheath. For this purpose, analytical expressions for glutamate diffusion in the glomerulus were considered in models with two-, three-, and fractional two-three-dimensional (2D-3D) geometry with an absorbing boundary. The time course of average glutamate concentration in the synaptic cleft of the mossy fiber-granule cell connection was calculated for both direct release of glutamate from the same synaptic unit, and for cumulative spillover of glutamate from neighboring release sites. Several kinetic schemes were examined, and the parameters of the diffusion models were estimated by identifying theoretical activation of AMPA receptors with direct release and spillover components of published experimental AMPA receptor-mediated EPSCs. For model selection, the correspondence of simulated paired-pulse ratio and EPSC increase after prevention of desensitization to experimental values were also taken into consideration. Our results suggest at least a 7- to 10-fold lower apparent diffusion coefficient of glutamate in the porous medium of the glomerulus than in water. The modeling of glutamate diffusion in the 2D-3D geometry gives the best fit of experimental EPSCs. We show that it could be only partly explained by normal diffusion of glutamate in the complex geometry of the glomerulus. We assume that anomalous diffusion of glutamate occurs in the glomerulus. A good match of experimental estimations and theoretical parameters, obtained in the simulations that use an approximation of anomalous diffusion by a solution for fractional Brownian motion, confirms our assumption.

Cerebellar Cortex↗
PubMed2006

Bayesian mapping of genotype x expression interactions in quantitative and qualitative traits.

A novel Bayesian gene mapping method, which can simultaneously utilize both molecular marker and gene expression data, is introduced. The approach enables a quantitative or qualitative phenotype to be expressed as a linear combination of the marker genotypes, gene expression levels, and possible genotype x gene expression interactions. The interaction data, given as marker-gene pairs, contains possible in cis and in trans effects obtained from earlier allelic expression studies, genetical genomics studies, biological hypotheses, or known pathways. The method is presented for an inbred line cross design and can be easily generalized to handle other types of populations and designs. The model selection is based on the use of effect-specific variance components combined with Jeffreys' non-informative prior--the method operates by adaptively shrinking marker, expression, and interaction effects toward zero so that non-negligible effects are expected to occur only at very few positions. The estimation of the model parameters and the handling of missing genotype or expression data is performed via Markov chain Monte Carlo sampling. The potential of the method including heritability estimation is presented using simulated examples and novel summary statistics. The method is also applied to a real yeast data set with known pathways.

Bayes Theorem↗
PubMed2006

GARD: a genetic algorithm for recombination detection.

MOTIVATION: Phylogenetic and evolutionary inference can be severely misled if recombination is not accounted for, hence screening for it should be an essential component of nearly every comparative study. The evolution of recombinant sequences can not be properly explained by a single phylogenetic tree, but several phylogenies may be used to correctly model the evolution of non-recombinant fragments. RESULTS: We developed a likelihood-based model selection procedure that uses a genetic algorithm to search multiple sequence alignments for evidence of recombination breakpoints and identify putative recombinant sequences. GARD is an extensible and intuitive method that can be run efficiently in parallel. Extensive simulation studies show that the method nearly always outperforms other available tools, both in terms of power and accuracy and that the use of GARD to screen sequences for recombination ensures good statistical properties for methods aimed at detecting positive selection. AVAILABILITY: Freely available http://www.datamonkey.org/GARD/

Algorithms↗
PubMed2003

Optimized bioregenerative space diet selection with crew choice.

Previous studies on optimization of crew diets have not accounted for choice. A diet selection model with crew choice was developed. Scenario analyses were conducted to assess the feasibility and cost of certain crew preferences, such as preferences for numerous-desserts, high-salt, and high-acceptability foods. For comparison purposes, a no-choice and a random-choice scenario were considered. The model was found to be feasible in terms of food variety and overall costs. The numerous-desserts, high-acceptability, and random-choice scenarios all resulted in feasible solutions costing between 13.2 and 17.3 kg ESM/person-day. Only the high-sodium scenario yielded an infeasible solution. This occurred when the foods highest in salt content were selected for the crew-choice portion of the diet. This infeasibility can be avoided by limiting the total sodium content in the crew-choice portion of the diet. Cost savings were found by reducing food variety in scenarios where the preference bias strongly affected nutritional content.

Astronauts↗
PubMed1993

NMR studies of 1H NOEs in glycogen.

We have examined the cross-relaxation behavior among the protons of oyster glycogen using nuclear Overhauser enhancement (NOE). Steady-state and transient NOEs were generated using low-power CW irradiation and frequency-selective inversions. In D2O, saturation of glycogen H2 and H4' at 3.64 ppm gave a strong negative NOE (eta = -0.74) at H1. The NOE was similar to the value predicted by the correlation time (tau c) calculated from the T1 and T2 of glycogen H1 in D2O assuming an isotropic rigid motor dipole-dipole model. Selective inversion of H2 and H4' gave a transient NOE at H1. In D2O, selective inversion of H1 also led to negative transient NOEs in the H2 + H4', H3, and H5 resonances. The magnitude and rates of appearance of the NOEs in H3 and H5 were too large to arise from direct H1-H3 and H1-H5 dipolar interactions, but were consistent with very efficient cross-relaxation leading to large second-order NOEs. The glycogen H1 NOE in H2O was also studied. Replacement of D2O with H2O as solvent significantly reduced the steady-state NOE at H1 following saturation of H2 + H4'. Saturation of the water resonance caused a large negative NOE at H1 (eta = -0.55) consistent with our earlier study which indicated that there was no direct dipolar interaction between H1 and free H2O.

Animals↗
PubMed1993

Lipid data from NHLBI veteran twins: interpreting genetic analyses when model assumptions fail.

Analyses were performed on lipid data from the NHLBI Veteran Twin Study. The analyses focused on longitudinal multivariate models, describing how the genetic effects on lipids vary over time. Our pedigree-based model selection approach allows simultaneous estimation of both covariance structure parameters and regression parameters. The analyses reveal strong correlations between additive genetic effects over time, implying that genetic effects on lipids are somewhat constant throughout the life span represented within this sample. Both univariate preliminary analyses and robust fitting applied to the longitudinal models indicate that several assumptions underlying the twin analyses are violated. Although variance component and correlation parameter estimates are not much changed by robust fitting analyses, questions remain about the behavior of parameter estimates in multivariate genetic models under departures from model assumptions.

Humans↗