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 991 records · Page 55Linked to original sources
PubMed1998

A generalized F mixture model for cure rate estimation.

Cure rate estimation is an important issue in clinical trials for diseases such as lymphoma and breast cancer and mixture models are the main statistical methods. In the last decade, mixture models under different distributions, such as exponential, Weibull, log-normal and Gompertz, have been discussed and used. However, these models involve stronger distributional assumptions than is desirable and inferences may not be robust to departures from these assumptions. In this paper, a mixture model is proposed using the generalized F distribution family. Although this family is seldom used because of computational difficulties, it has the advantage of being very flexible and including many commonly used distributions as special cases. The generalised F mixture model can relax the usual stronger distributional assumptions and allow the analyst to uncover structure in the data that might otherwise have been missed. This is illustrated by fitting the model to data from large-scale clinical trials with long follow-up of lymphoma patients. Computational problems with the model and model selection methods are discussed. Comparison of maximum likelihood estimates with those obtained from mixture models under other distributions are included.

Confidence Intervals↗
PubMed1999

On Bayesian calculations for mixture likelihoods and priors.

We present methodology for calculating Bayes factors between models as well as posterior probabilities of the models when the indicator variables of the models are integrated out of the posterior before Markov chain Monte Carlo (MCMC) computations. Standard methodology would include the indicator functions as part of the MCMC computations. We demonstrate that our methodology can give substantially greater accuracy than the traditional approach. We illustrate the methodology using the model selection prior of George and McCulloch applied to logistic regression and to a mixture model for observations in a hierarchical random effects model.

Bayes Theorem↗
PubMed2004

Marginalized transition models for longitudinal binary data with ignorable and non-ignorable drop-out.

We extend the marginalized transition model of Heagerty to accommodate non-ignorable monotone drop-out. Using a selection model, weakly identified drop-out parameters are held constant and their effects evaluated through sensitivity analysis. For data missing at random (MAR), efficiency of inverse probability of censoring weighted generalized estimating equations (IPCW-GEE) is as low as 40 per cent compared to a likelihood-based marginalized transition model (MTM) with comparable modelling burden. MTM and IPCW-GEE regression parameters both display misspecification bias for MAR and non-ignorable missing data, and both reduce bias noticeably by improving model fit.

Antipsychotic Agents↗
PubMed2002

CD34(+) or CD34(-): which is the more primitive?

Remarkable progress has been achieved in the characterization and isolation of primitive hematopoietic stem cells (HSC). HSC represent a very small subset of hematopoietic cells and provide self-renewal, possess differentiation capacity and allow a constant supply of the entire hematopoietic cell spectrum. Until recently, CD34 has been used as a convenient marker for HSC, since CD34(+) cells have been shown to possess colony-forming potential in short-term assays, maintain long-term colony-forming potential in in vitro cultures and allow the expression and differentiation of blood cells from different hematopoietic lineages in in vivo models. Clinical and experimental protocols have targeted CD34(+) cells enriched by a variety of selection models and have readily used these for transplantation, purging and gene therapies and targets for future organ replacement. Recent studies in murine and human models, however, have indicated that CD34(-) HSC exist as well, which possess engraftment potential and distinct HSC characteristics. These studies challenge the dogma that HSC are uniformly found in the CD34(+) subset, and question whether primitive HSC are CD34(+) or CD34(-). In this review, results on murine and human CD34(+) and CD34(-) HSC, differences between them and their possible interactions are examined.

Animals↗
PubMed2004

Migration of model contaminants from PET bottles: influence of temperature, food simulant and functional barrier.

To simulate post-consumer recycled plastics, selected model contaminants were incorporated into PET bottles using a time saving method. Migration into 3% acetic acid, a cola-type beverage and 95% ethanol was followed during 1 year of storage at 20 and 40 degrees C. Aroma compounds previously found in post-consumer PET material were used as model contaminants. Benzaldehyde was found to migrate to the highest extent. Storage at 40 degrees C affected the bottle material and this might be one reason for the high migration values of these bottles. Migration into ethanol was up to 20 times higher than into 3% acetic acid or a cola-type beverage. Bottles with a functional barrier resisted migration into food simulants even when filled with 95% ethanol and stored for 1 year at 40 degrees C. Differential scanning calorimetry measurements showed that ethanol was interacting with the plastic material. This resulted in a lower glass transition temperature of bottles stored with ethanol compared with bottles stored empty or with other food simulants.

Acetic Acid↗
PubMed2003

Genetic analysis of somatic cell score in Norwegian cattle using random regression test-day models.

The dataset used in this analysis contained a total of 341,736 test-day observations of somatic cell scores from 77,110 primiparous daughters of 1965 Norwegian Cattle sires. Initial analyses, using simple random regression models without genetic effects, indicated that use of homogeneous residual variance was appropriate. Further analyses were carried out by use of a repeatability model and 12 random regression sire models. Legendre polynomials of varying order were used to model both permanent environmental and sire effects, as did the Wilmink function, the Lidauer-Mäntysaari function, and the Ali-Schaeffer function. For all these models, heritability estimates were lowest at the beginning (0.05 to 0.07) and higher at the end (0.09 to 0.12) of lactation. Genetic correlations between somatic cell scores early and late in lactation were moderate to high (0.38 to 0.71), whereas genetic correlations for adjacent DIM were near unity. Models were compared based on likelihood ratio tests, Bayesian information criterion, Akaike information criterion, residual variance, and predictive ability. Based on prediction of randomly excluded observations, models with 4 coefficients for permanent environmental effect were preferred over simpler models. More highly parameterized models did not substantially increase predictive ability. Evaluation of the different model selection criteria indicated that a reduced order of fit for sire effects was desireable. Models with zeroth- or first-order of fit for sire effects and higher order of fit for permanent environmental effects probably underestimated sire variance. The chosen model had Legendre polynomials with 3 coefficients for sire, and 4 coefficients for permanent environmental effects. For this model, trajectories of sire variance and heritability were similar assuming either homogeneous or heterogeneous residual variance structure.

Animals↗
PubMedDate not supplied

Guidelines for selecting a conceptual model of nursing: focus on the individual patient.

How do conceptual models help you in clinical practice? How can you select the best model to guide successful interventions for your critical care patients? These authors present three guidelines for selection of a conceptual model for your patient situation. Six case studies of patients requiring critical care nursing demonstrate the fit of the patient's problems and goals for care with particular conceptual models of nursing.

Adult↗
PubMed2002

A SWOT analysis of the organization and financing of the Danish health care system.

The organization and financing of the Danish health care system was evaluated within a framework of a SWOT analysis (analysis of Strengths, Weaknesses, Opportunities and Threats) by a panel of five members with a background in health economics. The present paper describes the methods and materials used for the evaluation: selection of panel members, structure of the evaluation task according to the health care triangle model, selection of background material consisting of documents and literature on the Danish health care system, and a 1-week study visit.

Delivery of Health Care↗
PubMed1990

Choice and use of tracers.

The isotope dilution method has been widely employed to measure glucose turnover in man. The validity of this technique depends upon several assumptions. First, it is assumed that the selected model of glucose kinetics is valid. Under nonsteady state conditions this has recently been challenged for the most commonly used single compartment model. Secondly, it is assumed that the isotope is metabolized in the same manner as unlabeled glucose. If the isotope loses its label when subject to a substrate cycle (for example, [2-3H]glucose in glucose to glucose 6-phosphate cycling), an overestimate of glucose disposal will result. In contrast, if isotope that has been cleared (for example [6-3H]glucose in hepatic glucogen) is recycled into the systemic circulation, an underestimation of glucose turnover will result. Thirdly, it is assumed that measurement of specific activity is accurate. It has recently been shown that many commercially available tritiated (but not 14C) isotopes contain radioactive nonglucose contaminants which have a slower clearance rate than glucose under conditions of high glucose turnover. This can result in overestimates of specific activity and underestimates of glucose appearance, leading to calculated negative hepatic glucose release. While this problem may be avoided with purified tritiated glucose tracers, a similar problem with stable isotopes of glucose remains to be determined. In summary, the choice of glucose isotope should reflect the probability of substrate cycling and isotope cycling in the experiment in question. Care should be taken to ensure purified isotopes are selected and that the model of glucose kinetics used is valid during the conditions to be studied.

Animals↗
PubMed2007

Random changepoint modelling of HIV immunologic responses.

We propose a changepoint model for the analysis of longitudinal CD4 T-cell counts for HIV infected subjects following highly active antiretroviral treatment. The profile of CD4 counts for each subject follows a simple, 'broken stick' changepoint model, with random subject-specific parameters, including the changepoint. The model accounts for baseline covariates. The longitudinal CD4 records are censored at the time of the subject going off-study-treatment. This is a potentially informative drop-out mechanism, which we address by modelling it jointly with the CD4 count outcome. The drop-out model incorporates terms from the CD4 model, including the changepoint. The estimation is done in a Bayesian framework, with implementation via Markov chain Monte Carlo methods in the WinBUGS software. Model selection using DIC indicates that the data support the complex random changepoint and informative censoring model.

Antiretroviral Therapy, Highly Active↗
PubMed2003

Chlorophyll a fluorescence rise induced by high light illumination of dark-adapted plant tissue studied by means of a model of photosystem II and considering photosystem II heterogeneity.

Chlorophyll a fluorescence rise (FLR) measured in vivo in dark-adapted plant tissue immediately after the onset of high light continuous illumination shows complex O-K-J-I-P transient. The steps typically appear at about 400 micros (K), 2 ms (J), 30 ms (I), and 200 - 500 ms (P) and a transient decrease of fluorescence to local minima (dips D) can be observed after the K, J, and I steps. As the FLR reflects a function of photosystem II (PSII) and to more understand the FLR, a PSII reactions model was formulated comprising equilibrium of excited states among all light harvesting and reaction centre pigments and P680, reversible radical pair formation and the donor and acceptor side functions. Such a formulated model is the most detailed and complex model of PSII reactions used so far for simulations of the FLR. By varying of selected model parameters (rate constants and initial conditions) several conclusions can be made as for the origin of and changes in shape of the theoretical FLR and compare them with in-literature-reported results. For homogeneous population of PSII and using standard in-literature-reported values of the model parameters, the simulated FLR is characterized by reaching the minimal fluorescence F(0) at about 3 ns after the illumination is switched on lasting to about 1 micros, followed by fluorescence rise to a plateau located at about 2 ms and subsequent fluorescence rise to a global maximum that is reached at about 60 ms. Varying of the values of rate constants of fast processes that can compete for utilization of the excited states with fluorescence emission does not change qualitatively the shape of the FLR. However, primary photochemistry of PSII (the charge separation, recombination and stabilization), non-radiative loss of excited states in light harvesting antennae and excited states quenching by oxidized plastoquisnone (PQ) molecules from the PQ pool seem to be the main factors controlling the maximum quantum yield of PSII photochemistry as expressed by the F(V)/F(M) ratio. The appearance of the plateau at about 2 ms in the FLR is affected by several factors: the height of the plateau in the FLR increases when the fluorescence quenching by oxidized P680(+) is not considered in the simulations or when the electron transfer from Q(A)(-) to Q(B)((-)) is slowed down whereas the height of the plateau decreases and its position is shifted to shorter times when OEC is initially in higher S state. The plateau at about 2 ms is changed into the local fluorescence maximum followed by a dip when the fluorescence quenching by oxidized PQ molecules or the charge recombination between P680(+) and Q(A)(-) is not considered in the simulations or when all OEC is initially in the S(0) state or when the S -state transitions of OEC are slowed down. Slowing down of the S -state transitions of OEC as well as of the electron transfer from Q(A)(-) to Q(B)((-)) also causes a decrease of maximal fluorescence level. In the case of full inhibition of the S -state transitions of OEC as well as in the case of full inhibition of the electron donation to P680(+) by Y(Z), the local fluorescence maximum becomes the global fluorescence maximum. Assuming homogeneous PSII population, theoretical FLR curve that only far resembles experimentally measured O-J-I-P transient at room temperature can be simulated when slowly reducing PQ pool is considered. Assuming heterogeneous PSII population (i.e. the alpha/beta and the Q(B) -reducing/Q(B)-non-reducing heterogeneity and heterogeneity in size of the PQ pool and rate of its reduction) enables to simulate the FLR with two steps between minimal and maximal fluorescence whose relative heights are in agreement with the experiments but not their time positions. A cause of this discrepancy is discussed as well as different approaches to the definition of fluorescence signal during the FLR.

Chlorophyll↗
PubMed2002

Development of a genetic selection for catalytic antibodies.

The design and evaluation of a new genetic selection system for evolving catalytic antibodies with aldolase activity are described. Through a series of model selections, we have identified selection conditions where expression of a catalytically active antibody confers a growth advantage to Escherichia coli. In addition, we provide evidence that the growth advantage is a direct result of catalytic activity.

Antibodies, Catalytic↗
PubMed1993

Species selection on variability.

Most analyses of species selection require emergent, as opposed to aggregate, characters at the species level. This "emergent character" approach tends to focus on the search for adaptations at the species level. Such an approach seems to banish the most potent evolutionary property of populations--variability itself--from arguments about species selection (for variation is an aggregate character). We wish, instead, to extend the legitimate domain of species selection to aggregate characters. This extension of selection theory to the species level will concentrate, instead, on the relation between fitness and the species character, whether aggregate or emergent. Examination of the role of genetic variability in the long-term evolution of clades illustrates the cogency of broadening the definition of species selection to include aggregate characters. We reinterpret, in this light, a classic case presented in support of species selection. As originally presented, the species selection explanation of volutid neogastropod evolution was vulnerable to a counterinterpretation at the organism level. Once this case is recast within a definition of species selection that reflects the essential structure and broad applicability of hierarchical selection models, the organism-level reinterpretation of variability loses its force. We conclude that species selection on variability is a major force of macroevolution.

Journal Article↗
PubMed2005

The SWISS-MODEL workspace: a web-based environment for protein structure homology modelling.

MOTIVATION: Homology models of proteins are of great interest for planning and analysing biological experiments when no experimental three-dimensional structures are available. Building homology models requires specialized programs and up-to-date sequence and structural databases. Integrating all required tools, programs and databases into a single web-based workspace facilitates access to homology modelling from a computer with web connection without the need of downloading and installing large program packages and databases. RESULTS: SWISS-MODEL workspace is a web-based integrated service dedicated to protein structure homology modelling. It assists and guides the user in building protein homology models at different levels of complexity. A personal working environment is provided for each user where several modelling projects can be carried out in parallel. Protein sequence and structure databases necessary for modelling are accessible from the workspace and are updated in regular intervals. Tools for template selection, model building and structure quality evaluation can be invoked from within the workspace. Workflow and usage of the workspace are illustrated by modelling human Cyclin A1 and human Transmembrane Protease 3. AVAILABILITY: The SWISS-MODEL workspace can be accessed freely at http://swissmodel.expasy.org/workspace/

Algorithms↗
PubMed1996

Analysing the relationship between maternal weight gain and birthweight: exploration of four statistical issues.

Four statistical issues concerning the analysis of birthweight and maternal weight gain during pregnancy are discussed: (1) Part-whole correlation is described (e.g. the correlation between total maternal weight gain and her infant's birthweight). (2) The choice between a ratio or two separate explanatory variables is explored (e.g. body mass index or using maternal weight and height separately). (3) Two statistical properties (bias and power) when a binary variable replaces a continuous variable are discussed (e.g. consequences of using low birthweight instead of reported birthweight). (4) A model selection procedure is presented to provide a way to select a useful subset of variables from a large number of available explanatory variables to model an outcome variable (e.g. birthweight). These issues are illustrated with a set of 4017 births from Moffitt Hospital at the University of California, San Francisco. Furthermore, these four issues arise in a number of applications of statistical methods to data collected to study the epidemiology of newborn infants.

Bias↗
PubMed1994

Semiparametric models for antagonistic drug interactions.

A new class of models to describe antagonistic drug interactions are presented. They are semiparametric in that they use nonparametric functions (splines) but are forced to obey certain constraints corresponding to reasonable assumptions. We propose the models primarily for exploratory data analysis, but they may also be definitive models for such purposes as predicting future responses. Certain problems that arise in semiparametric modeling, such as model selection, are addressed so that we can propose a relatively automatic and objective approach to model determination. We demonstrate the applicability of the class of models we propose to two real data set examples involving pain relief response to opioid agonists/antagonists. The results suggest that the semiparametric approach is particularly useful when unusual shapes link dose to response.

Animals↗
PubMed2001

Disposition of sulfadimethoxine in camels (Camelus dromedarius) following single intravenous and oral doses.

Single-dose pharmacokinetics of sulfadimethoxine were determined in six adult camels (Camelus dromedarius) following administration of a mean dosage of 17.5 +/- 2.7 mg/kg both i.v. and p.o. Serial blood samples were collected through an indwelling jugular catheter intermittently for 5 days for both routes. Sulfadimethoxine was assayed using high-performance liquid chromatography. Serum drug concentration versus time data for each animal was subjected to linear regression, with the best-fit model selected based on residual analysis. The data fit best into a two-compartment open model, with first-order input for oral administration. For orally administered drug, mean maximum serum concentration of 19.3 +/- 1.7 microg/ml was reached at 11.41 +/- 2.59 hr, with an elimination rate constant of 0.09/hr +/- 0.05/hr and an elimination half-life of 11.7 +/- 3 hr. Mean peak serum concentration following i.v. administration was 223 +/- 48 microg/ml. Mean volume of distribution at steady state was 0.393 +/- 0.049 L/kg. Elimination rate constants differed with i.v. and oral administration, suggesting a flip-flop model. Oral bioavailability was 103% +/- 38%. Comparison of maximum serum concentrations to the microbial breakpoint concentration reported for sulfadimethoxine (512 microg/ml) suggests that the dose used in this study, 17.5 +/- 2.7 mg/kg, is insufficient for achieving therapeutic serum levels.

Administration, Oral↗
PubMed2000

Comparison of models for describing the lactation curve of latxa sheep and an analysis of factors affecting milk yield.

The objectives of this work were 1) to compare the goodness-of-fit of empirical models of the lactation curve and 2) to analyze the factors affecting the shapes of the lactation curves, the parameters describing them, and the overall milk yield of Latxa dairy sheep. A total of 14,699 records from 2711 ewes, collected during three consecutive years (1995 to 1997) by the milk recording program of the Latxa ewe of the Basque Country (Spain), were used. Six mathematical models and three fitting procedures were compared. The estimation of model parameters by nonlinear fitting procedures was superior to that by linear regression methods. A nonlinear variable decay model fitted the data better than the other models, as judged by lower mean square prediction error, residual sums of squares, and a lack of first-order positive autocorrelation as assessed by the Durbin Watson coefficient. The effects of the flock, flock-year interaction, month of lambing, length of lactation nested within month of lambing, parity, and number of live lambs born had significant effects on the parameters of the model and the total milk yield (P < 0.01). The prediction of milk yield from the selected model was similar to the estimates obtained with the Fleischmann method currently used by the national breeding program for the Latxa breed.

Animals↗