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 235 records · Page 13Linked to original sources
PubMed2007

Bookmark locations and item response model selection in the presence of local item dependence.

The bookmark standard setting procedure is a popular method for setting performance standards on state assessment programs. This study reanalyzed data from an application of the bookmark procedure to a passage-based test that used the Rasch model to create the item ordered booklet. Several problems were noted in this implementation of the bookmark procedure, including disagreement among the SMEs about the correct order of items in the bookmark booklet, performance level descriptions of the passing standard being based on passage difficulty as well as item difficulty, and the presence of local item dependence within reading passages. Bookmark item locations were recalculated for the IRT three-parameter model and the multidimensional bifactor model. The results showed that the order of item locations was very similar for all three models when items of high difficulty and low discrimination were excluded. However, the items whose positions were the most discrepant between models were not the items that the SMEs disagreed about the most in the original standard setting. The choice of latent trait model did not address problems of item order disagreement. Implications for the use of the bookmark method in the presence of local item dependence are discussed.

Achievement↗
PubMed2003

Translating multiple assessment techniques into an intervention selection model for classrooms.

Translating current research to school-based clinical practice highlights issues not often encountered in laboratory settings. With the assistance of a consultant, teachers conducted functional analyses, brief multielement treatment comparisons, and controlled treatment evaluations under naturalistic conditions in the classroom. Teachers also provided input on treatment selection. Treatment integrity data collected throughout the study suggested that teachers implemented analyses and treatments with high integrity. The functional analysis outcomes combined with effectiveness and acceptability data led to the selection of interventions that reduced problem behavior in the classrooms for each of 3 children.

Behavior Therapy↗
PubMed1978

Physiological perfusion model for cephalosporin antibiotics I: Model selection based on blood drug concentrations.

Various cephalosporins with different degrees of protein binding were administered to human volunteers. Blood samples were collected as a function of time and were assayed for drug content by a microbiological assay. A pharmacokinetic analysis of the data was performed using a two-compartment model with and without protein binding in the central compartment and a perfusion model. Both the two-compartment model without protein binding and the physiological perfusion model adequately described the blood levels of all three cephalosporins.

Blood Proteins↗
PubMed1991

Identification of insulin receptor systems: assessing the impact of model selection and measurement error on precision of parameter estimates using Monte Carlo study.

An extensive Monte Carlo study has been carried out in order to study the effect of measurement error on the precision of parameter estimates of an insulin binding system. Hypothetical radioimmunoassay experiments were generated for insulin binding to erythrocytes. The design of experiments followed strictly the protocol of real experiments. Randomly generated error was added to the synthetic data. The standard technique, a weighted non-linear regression analysis, was employed to re-estimate parameters of a model of two receptor sites and a model of negative co-operativity. As the original parameter values were known, the differences between original and estimated values was studied for (a) measurement error in the range from 0-17%, (b) random initial estimates and (c) error-free non-specific binding. In addition, analytical estimates of parameter precision were compared with the true between-experiment variation of parameter estimates. At the measurement error of 12%, a one site model is recommended to estimate the high affinity population of the two sites model. Plausible results can be expected in 90% of experiments, the between-experiment variation being approximately 30%. The model of two receptor sites gives approximately two thirds of plausible results. The high affinity population can be estimated with the between-experiment variation of 40%, the low affinity population is virtually unidentifiable with the between-experiment variation of approximately 100% and parameter estimates biased to higher values. Only half of the results obtained from the model of negative co-operativity are plausible, the variation in parameter estimates ranges from 90-150% and estimates are biased to higher values. At the level of 12% measurement error, random initial estimates do not significantly affect the estimation process, provided initial estimates are selected from a feasible range. At the same measurement error, the error-free non-specific binding does not improve the results, indicating that the mean of six replicates may be taken as a reliable estimate of non-specific binding. The analytical estimates of the coefficient of variation systematically underestimates the true between-experiments coefficient of variation, the difference has been found to be about 50%.

Erythrocytes↗
PubMed2003

A method for evaluating the results of Bayesian model selection: application to linkage analyses of attributes determined by two or more genes.

OBJECTIVES: We apply and evaluate the intrinsic Bayes factor (IBF) of Berger and Pericchi [J Am Stat Assoc 1996;91:109-122; Bayesian Statistics, Oxford University Press, vol 5, 1996] to linkage analyses done using the stochastic search variable selection (SSVS) method of George and McCulloch [J Am Stat Assoc 1993;88:881-889] as proposed by Suh et al. [Genet Epidemiol 2001;21(suppl 1):S706-S711]. METHODS: We consider 20 simulations of linkage data obtained under two different generating models. The SSVS is applied to a multiple regression extension [Genet Epidemiol 2001;21(suppl 1): S706-S711] of the Haseman-Elston [Behav Genet 1972;2:3-19; Genet Epidemiol 2000;19:1-17] methods. Four prior distributions are considered. We apply the IBF criterion to those samples where different prior distributions result in different top models. RESULTS: In those samples where three different models were obtained using the four priors, application of the IBFs eliminated one of the two wrong models in 4 out of 5 situations. Further elimination using the IBF criterion for situations with two different subsets did not serve as well. CONCLUSIONS: When different priors result in three or more different subsets of markers, one can use the IBF to get this number down to two for consideration. When two subsets result we recommend that both be considered.

Bayes Theorem↗
PubMed1987

Model selection for ventricular mechanics: a sensitivity analysis approach.

Quantitative characterization of left ventricle pump properties has been recognized as being of great significance for both physiological and clinical purposes. Several descriptions have been proposed in the past to this end, where the ventricle is viewed as an isovolumic pressure generator coupled to an internal impedance, considered as either only viscous, only elastic or viscoelastic. Though these models have been used widely, the respective advantages and limits have not been fully elucidated. In this paper, six models for the left ventricular pumping function, of the viscoelastic type, are compared using both simulated and experimental data in a typical parameter estimation approach. Elastic and viscous parameters are estimated starting from ventricular pressure and aortic flow, together with the isovolumic pressure at the same preload. The basis for the comparison is the well-established criterion relating the fit obtained from collected data and the covariance matrix of the parameter estimates. The latter allows evaluation of the so-called indifference region in the parameter space, which is represented by an ellipse if both elastic and viscous elements are present. The properties of the indifference region are synthetically represented by two indices linked to the area and the eccentricity of the ellipse: the first represents the mean accuracy of the parameter estimate, the second gives information about the different sensitivities to variation of single parameters. This comparison, in both simulated and experimental cases, generally leads to preference for a model where elastance and viscosity vary with time in linear proportion to the isovolumically developed ventricular pressure. Appropriate description of the elastic effect reveals it to be very crucial while the viscous effect, though improving the fitting of data, is less critical.

Animals↗
PubMed2006

Are fMRI event-related response constant in time? A model selection answer.

An accurate estimation of the hemodynamic response function (HRF) in functional magnetic resonance imaging (fMRI) is crucial for a precise spatial and temporal estimate of the underlying neuronal processes. Recent works have proposed non-parametric estimation of the HRF under the hypotheses of linearity and stationarity in time. Biological literature suggests, however, that response magnitude may vary with attention or ongoing activity. We therefore test a more flexible model that allows for the variation of the magnitude of the HRF with time in a maximum likelihood framework. Under this model, the magnitude of the HRF evoked by a single event may vary across occurrences of the same type of event. This model is tested against a simpler model with a fixed magnitude using information theory. We develop a standard EM algorithm to identify the event magnitudes and the HRF. We test this hypothesis on a series of 32 regions (4 ROIS on eight subjects) of interest and find that the more flexible model is better than the usual model in most cases. The important implications for the analysis of fMRI time series for event-related neuroimaging experiments are discussed.

Algorithms↗
PubMed2005

3D-extravasation model -- selection of highly motile and metastatic cancer cells.

Extravasation has been described as a rate-limiting step in the process of hematogeneous metastasis formation. Thereby, transendothelial migration of tumor cells consists of a complex series of events involving multiple cell-cell and cell-matrix interactions. 3D-extravasation assays are valuable tools for the identification of genes, which are the key players at switchboards of the intracellular signaling pathways. In consequence, the combination of 3D-modeling and whole genome expression analysis lead to unravel molecular parameters which descripe distinct clinical phenotypes of cancer and therefore, work as prognosticators, predictors of therapy and new therapy targets.

Cell Movement↗
PubMed1999

Development of a Bayesian Network for the prognosis of head injuries using graphical model selection techniques.

The assessment of a head-injured patient's prognosis is a task that involves the evaluation of diverse sources of information. In this study we propose an analytical approach, using a Bayesian Network (BN), of combining the available evidence. The BN's structure and parameters are derived by learning techniques applied to a database (600 records) of seven clinical and laboratory findings. The BN produces quantitative estimations of the prognosis after 24 hours for head-injured patients in the outpatients department. Alternative models are compared and their performance is tested against the success rate of an expert neurosurgeon.

Bayes Theorem↗
PubMed2001

Lymphatic transport of proteins after s.c. injection: implications of animal model selection.

Subcutaneous (s.c.) administration continues to be the main route for the delivery of protein drugs due to their poor bioavailability by most non-parenteral routes. While small drug molecules are rapidly and extensively absorbed after s.c. injection, the systemic bioavailability of protein drugs is often incomplete and variable. Given the widespread use of the s.c. route for protein drugs, surprisingly little is known about the factors that govern the rate and extent of protein absorption from the interstitial space and the role of the lymphatic system in the transport of these molecules to the systemic circulation. The few studies that have directly addressed the role of lymphatic transport in protein bioavailability are complicated by the use of methods and models that vary widely. In this review we will evaluate the available literature describing the lymphatic transport of proteins after s.c. injection and more specifically, address the impact of experimental variation (e.g. site of cannulation, animal model, anesthesia) on the interpretation of the data obtained. We will also describe in some detail the sheep model currently in use in our laboratory, which allows both estimation of the extent of uptake of protein drugs into the lymphatics draining the injection site, and quantification of the contribution of lymphatic transport to the absolute bioavailability.

Animals↗
PubMed2004

Assessing the quality of molecular divergence time estimates by fossil calibrations and fossil-based model selection.

Estimates of species divergence times using DNA sequence data are playing an increasingly important role in studies of evolution, ecology and biogeography. Most work has centred on obtaining appropriate kinds of data and developing optimal estimation procedures, whereas somewhat less attention has focused on the calibration of divergences using fossils. Case studies with multiple fossil calibration points provide important opportunities to examine the divergence time estimation problem in new ways. We discuss two cross-validation procedures that address different aspects of inference in divergence time estimation. 'Fossil cross-validation' is a procedure used to identify the impact of different individual calibrations on overall estimation. This can identify fossils that have an exceptionally large error effect and may warrant further scrutiny. 'Fossil-based model cross-validation' is an entirely different procedure that uses fossils to identify the optimal model of molecular evolution in the context of rate smoothing or other inference methods. Both procedures were applied to two recent studies: an analysis of monocot angiosperms with eight fossil calibrations and an analysis of placental mammals with nine fossil calibrations. In each case, fossil calibrations could be ranked from most to least influential, and in one of the two studies, the fossils provided decisive evidence about the optimal molecular evolutionary model.

Animals↗
PubMed2026

Multi-omics analysis identifies key genes and functional loci affecting teat number in American Large White and Landrace pigs and their application in optimizing genomic selection models.

BACKGROUND: Teat number is a crucial economic trait in pigs. It directly affects the ability of sows to lactate, which in turn influences the survival and health of piglets. The teat number of French Large White pigs is close to 16, while the teat number of American Large White and Landrace pigs is about 14. In order to improve the teat number of American Landrace and Large White pigs through molecular approaches and precise breeding techniques, we genotyped 2,131 American Landrace and 4,564 American Large White with teat number phenotype using a 50 K SNP chip. Then, the SNP-chip data was imputed to the level of whole-genome sequencing (iWGS). Based on iWGS data, we conducted GWAS to identify novel, significant SNPs associated with teat number and to incorporate them into genomic selection. RESULTS: In Landrace pigs, significant SNPs for TTN mapped to SSC2, SSC7, SSC8, and SSC14; the SSC8 and SSC14 effects are novel. LTN mapped to SSC7, RTN to SSC7 and SSC8. The lead SSC7 SNP explained 2.60% of TTN phenotypic variance. In Large White pigs, significant SNPs were detected on SSC7 and SSC10 for TTN; SSC7, SSC10, and SSC12 for LTN; and SSC7 and SSC10 for RTN. The most significant locus on SSC7 accounted for 2.99% of the phenotypic variance in TTN. Additionally, a multi-population meta-analysis detected significant novel SNPs for LTN on SSC1 and SSC8. By utilizing Bayesian fine mapping, the most precise QTL confidence interval on SSC7 for both TTN and RTN in Large White pigs was reduced to 40 kb. By integrating functional gene annotation with RNA-seq and ATAC-seq data from Erhualian and Bamaxiang pigs mammary placodes at embryonic day 26, we prioritized PTPN13, TRPV3, ZDHHC13, and BRD2 as novel candidate genes for teat number. We then incorporated the significant SNPs to GBLUP and benchmarked genomic-selection accuracy. In both breeds, fitting the top SNP as fixed maximized prediction for TTN and RTN, whereas treating all significant loci as an additional random effect optimized LTN. CONCLUSIONS: Our findings provide a theoretical basis for dissecting new key genes affecting teat number and for advancing molecular breeding of teat number in pigs.

Animals↗
PubMed1995

Pharmacokinetic model selection for target controlled infusions of propofol. Assessment of three parameter sets.

BACKGROUND: Computer-assisted target controlled infusions (TCI) result in prediction errors that are influenced by pharmacokinetic variability among and within patients. It is uncertain whether the selection of a propofol pharmacokinetic parameter set significantly influences drug concentrations and clinical acceptability. METHODS: Thirty patients received similar propofol TCI regimens after being randomly allocated to one of three parameter sets. Arterial and venous concentrations were measured and prediction errors calculated from pooled and intrasubject data. RESULTS: Arterial propofol concentrations in the Dyck group revealed greater bias (mean 43%) than did those in the Marsh (-1%) and Tackley (-3%) groups. The Dyck group also showed greater inaccuracy (mean:47%) than the Marsh (29%) and Tackley (24%) groups. There was little tendency for measured concentrations to vary from targeted values over time (divergence). Variability about an observed mean in individual patients (wobble) was low. Venous propofol concentrations were initially much less than arterial concentrations, but this difference decreased over time. CONCLUSIONS: Although it may be preferable to administer propofol TCI by using a locally derived parameter set, it is acceptable to use a model from elsewhere. The Marsh and Tackley models produced equally good performance and are appropriate for propofol TCI within the range of 3-6 micrograms/ml. The Dyck model was less accurate at maintaining anesthetic concentrations, possibly because it was derived from low concentrations. Concentrations in blood, the most sensitive indicators of performance, demonstrated differences among the parameter sets. Clinically, TCI worked well, and by clinical criteria, the choice of pharmacokinetic model did not appear to make a difference.

Adult↗
PubMed2006

Bias in error estimation when using cross-validation for model selection.

BACKGROUND: Cross-validation (CV) is an effective method for estimating the prediction error of a classifier. Some recent articles have proposed methods for optimizing classifiers by choosing classifier parameter values that minimize the CV error estimate. We have evaluated the validity of using the CV error estimate of the optimized classifier as an estimate of the true error expected on independent data. RESULTS: We used CV to optimize the classification parameters for two kinds of classifiers; Shrunken Centroids and Support Vector Machines (SVM). Random training datasets were created, with no difference in the distribution of the features between the two classes. Using these "null" datasets, we selected classifier parameter values that minimized the CV error estimate. 10-fold CV was used for Shrunken Centroids while Leave-One-Out-CV (LOOCV) was used for the SVM. Independent test data was created to estimate the true error. With "null" and "non null" (with differential expression between the classes) data, we also tested a nested CV procedure, where an inner CV loop is used to perform the tuning of the parameters while an outer CV is used to compute an estimate of the error. The CV error estimate for the classifier with the optimal parameters was found to be a substantially biased estimate of the true error that the classifier would incur on independent data. Even though there is no real difference between the two classes for the "null" datasets, the CV error estimate for the Shrunken Centroid with the optimal parameters was less than 30% on 18.5% of simulated training data-sets. For SVM with optimal parameters the estimated error rate was less than 30% on 38% of "null" data-sets. Performance of the optimized classifiers on the independent test set was no better than chance. The nested CV procedure reduces the bias considerably and gives an estimate of the error that is very close to that obtained on the independent testing set for both Shrunken Centroids and SVM classifiers for "null" and "non-null" data distributions. CONCLUSION: We show that using CV to compute an error estimate for a classifier that has itself been tuned using CV gives a significantly biased estimate of the true error. Proper use of CV for estimating true error of a classifier developed using a well defined algorithm requires that all steps of the algorithm, including classifier parameter tuning, be repeated in each CV loop. A nested CV procedure provides an almost unbiased estimate of the true error.

Algorithms↗
PubMed1990

Histocompatible miniature, boar model: selection of transformed cell lines of B and T lineages producing retrovirus.

A lymphoblastoid cell line (B1) was isolated in culture following a brief exposure to 5-azacytidine from peripheral-blood mononuclear cells of a boar previously injected with cells (Shimozuma) producing porcine retrovirus (Tsukuba-1) and suffering a severe non-neoplastic syndrome at autopsy. B1 cell line and 5 of its sublines were propagated for more than 100 generations, retaining doubling times comprised between 16.8 and 27.5 hr and growing readily in agarose or agar (plating efficiency: 5 to 50%). Karyotype analyses showed that 4 sublines were nearly diploid, except for cells of L14, which displayed a monosomy affecting chromosome 18 pair. Two sublines (L35 and L45) were considered as being of T-cell lineage, since MSA, antigen was observed on the surface of approximately 30% of cells. Three sublines (L23, L14 and L52) were considered of B-cell lineage, since membrane immunoglobulins were observed on the cell surface. In addition, sublines L23 and L52 were actively secreting immunoglobulin of mu isotype. Retrovirus particles were evidenced in gradient-purified preparation of 200-fold-concentrated cell culture supernatants of the B1 cell line, L14, L35 and L52 sublines, using both a reverse transcriptase activity assay and electron microscopic observation. These cell lines can be used to select for porcine retrovirus variants with transforming potential for lymphocytes of B and T lineages.

Animals↗
PubMed2000

Evolutionarily stable strategy in a sex- and frequency-dependent selection model.

In this paper, a sex-dependent matrix game haploid model is investigated. For this model, since the phenotypes of female and male individuals are determined by alleles located at a single locus and are sex dependent, any given genotype corresponds to a strategy pair. Thus, a strategy pair is an ESS if and only if the allele corresponding to this strategy pair cannot be invaded by any mutant allele. We show that an ESS equilibrium must be locally asymptotically stable if it exists.

Alleles↗