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 505 records · Page 28Linked to original sources
PubMed1989

Poor repertoire selection in symmetric idiotypic network models.

The selection of B and T cell repertoires is known to be influence by idiotypic interactions during early ontogeny. Early B cell clones are multi-specific, have numerous idiotypic interactions, produce IgM antibodies, and may constitute a separate cell lineage (characterised by the Ly1 or CD5 marker). Furthermore, because early B cells are self-reactive, self antigens should play a crucial role in the repertoire selection. Previously we developed a theoretical model of idiotypic B cell interactions. The model is based on a (symmetric) bell-shaped interaction function. The symmetry and the shape are a consequence of the process of receptor crosslinking. Assuming that early (i.e. IgM) B cell interactions are independent of helper T cell activation, we now apply this model to the problem of idiotypic repertoire selection. In the model the molecular structures of B cell receptors (i.e. of the idiotypes) and of self antigens are represented by random (bit)patterns. Interactions are based on complementary matches between these patterns. Therefore, the repertoire selection is brought about by stimulatory networks based upon complementary matches. These results show that the presence of a self antigen specifically shapes the B cell repertoire. The selection depends on the nature of the antigenic signal; we incorporate either stimulatory or inhibitory (i.e. tolerizing) self antigens. Clones stimulated by the self antigen tend to become aggressive in the network; tolerized self-reactive clones tend to become suppressed. Thus, idiotypic interactions play a facilitating role in self/non-self discrimination. However, since the idiotypic selection is only a tendency, the immune system should not rely on it. We next incorporate two classes of B cells. We call them the early, or "IgM", and the late, or "IgG", B cells (IgG B cells appear after the development of the IgM repertoire). IgG B cells may have only a few idiotypic interactions. We investigate whether the early IgM repertoire, which is (partly) selected by self antigen, influences the selection of the late IgG repertoire. It turns out that this influence can only be non-specific. Either we find a similar tendency in the IgG repertoire, or we find that the IgG repertoire influences the IgM repertoire non-specifically. Thus, despite the fact that this theoretical work readily confirms empirical results showing that the manipulation of early idiotypic interactions can have specific and long-lasting effects, the presupposed physiological role of these interactions, in terms of self-reactivity or repertoire selection, fails to develop in our models.

Animals↗
PubMed1980

Three-receptor, clonal expansion model for selection of self-recognition in the thymus.

The postulate is made that dual recognition by T lymphocytes is due to two types of receptor, one encoded by antibody genes and one by a distinct multigene family with simple rules for expression. This postulate leads to a model explaining ontogenic and evolutionary selection for self recognition, T lymphocyte effector function, the apparent high frequency of alloreactive T cells and immune response gene activity. The model is contrasted with previous explanations of self-recognition phenomena.

Animals↗
PubMed2004

Over-winter survival in subadult European rabbits: weather effects, density dependence, and the impact of individual characteristics.

The survival probability of an individual may be limited by density-dependent mechanisms and by environmental stochasticity, but can also be modified by individual characteristics. In our study, we investigated over-winter survival of subadults of an enclosed European rabbit Oryctolagus cuniculus population in a temperate zone habitat over the period 1992-2002. We: (1) selected for appropriate models to explain individual variation in over-winter survival and the animals' autumn body mass, the latter was used as a measure of the individual pre-winter body condition; and (2) aimed to compare the sensitivity of the target variables on the realised variation of the factors considered. Model selection based on information theory revealed that individual over-winter survival was best explained by the combination of autumn body mass, winter temperature, population density and sex, where the probability of survival was higher in females than in males. According to this model, the probability of survival reacted most sensitively to variation in the autumn body mass and in winter temperature. Individual autumn body mass was best explained by the combination of the date of birth, population density, and weather conditions by means of the percentage of rainy days during the first 2 months after the animals had emerged above ground, where the autumn body mass was negatively related to the percentage of rainy days. The chosen model suggested that the autumn body mass reacted most sensitively to variation in the date of birth. Combining these models, we found that weather conditions during two different periods of time as well as population density, sex and the date of birth operated together to determine the probability of over-winter survival. In particular, the study points out the high impact of environmental stochasticity on over-winter survival: (1) by direct effects of winter temperature conditions, and (2) by the indirect action of weather conditions to which the animals were exposed during the early period of juvenile development.

Age Factors↗
PubMed2005

Sulfoxidation mechanisms catalyzed by cytochrome P450 and horseradish peroxidase models: spin selection induced by the ligand.

The sulfoxidation of dimethyl sulfide (DMS), by two different heme-type enzyme models (without the protein), namely, horseradish peroxidase (HRP) and cytochrome P450 (P450), was studied using density functional theory. The models differ from each other by the axial ligand of the iron, which is imidazole in the case of HRP and thiolate in the case of P450. The computational results reveal a concerted oxygen atom transfer to sulfur, with spin-state selection dependent upon the identity of the proximal ligand. In the case of thiolate, the mechanism prefers the high-spin quartet pathway; whereas in the case of imidazole, the mechanism involves two-state reactivity (TSR), with competing quartet and doublet spin states. Furthermore, with thiolate the high-spin transition state, (4)TS(P450), has an upright conformation with a large Fe-O-S(DMS) angle of 147 degrees , whereas the low-spin species, (2)TS(P450), has a small angle and its Fe-O moiety makes an O-N(Por) bond with one of the nitrogen atoms of the porphine macrocycle. By contrast, when the proximal ligand is imidazole, both transition states possess a bent Fe-O bond and an O-N(Por) bond. These spin-state selection patterns obey simple orbital-selection rules, which are manifestations of the electronic nature of the ligand, i.e., the electron-releasing effect of the thiolate vis-a-vis the electron-withdrawal effect of imidazole. Other possible reactivity expressions of the spin-selection patterns are discussed [Dowers, T. S., Rock, D. A., Rock, D. A., Jones, J. P. (2004) J. Am. Chem. Soc. 126, 8868-8869]. Theory shows that intrinsically, HRP should be as reactive as P450 toward sulfoxidation.

Armoracia↗
PubMed2002

Summary of the National Toxicology Program's report of the endocrine disruptors low-dose peer review.

At the request of the U.S. Environmental Protection Agency (U.S. EPA), the National Toxicology Program organized an independent and open peer review to evaluate the scientific evidence on low-dose effects and nonmonotonic dose-response relationships for endocrine-disrupting chemicals in mammalian species. For this peer review, "low-dose effects" referred to biologic changes that occur in the range of human exposures or at doses lower than those typically used in the standard testing paradigm of the U.S. EPA for evaluating reproductive and developmental toxicity. The demonstration that an effect is adverse was not required because in many cases the long-term health consequences of altered endocrine function during development have not been fully characterized. A unique aspect of this peer review was the willing submission of individual animal data by principal investigators of primary research groups active in this field and the independent statistical reanalyses of selected parameters prior to the peer review meeting by a subpanel of statisticians. The expert peer-review panel (the panel) also considered mechanistic data that might influence the plausibility of low-dose effects and identified study design issues or other biologic factors that might account for differences in reported outcomes among studies. The panel found that low-dose effects, as defined for this review, have been demonstrated in laboratory animals exposed to certain endocrine-active agents. In some cases where low-dose effects have been reported, the findings have not been replicated. The shape of the dose-response curves for reported effects varied with the end point and dosing regimen and were low-dose linear, threshold-appearing, or nonmonotonic. The findings of the panel indicate that the current testing paradigm used for assessments of reproductive and developmental toxicity should be revisited to see whether changes are needed regarding dose selection, animal-model selection, age when animals are evaluated, and the end points being measured following exposure to endocrine-active agents.

Androgens↗
PubMed2003

A general model for selectivity in olefin cross metathesis.

In recent years, olefin cross metathesis (CM) has emerged as a powerful and convenient synthetic technique in organic chemistry; however, as a general synthetic method, CM has been limited by the lack of predictability in product selectivity and stereoselectivity. Investigations into olefin cross metathesis with several classes of olefins, including substituted and functionalized styrenes, secondary allylic alcohols, tertiary allylic alcohols, and olefins with alpha-quaternary centers, have led to a general model useful for the prediction of product selectivity and stereoselectivity in cross metathesis. As a general ranking of olefin reactivity in CM, olefins can be categorized by their relative abilities to undergo homodimerization via cross metathesis and the susceptibility of their homodimers toward secondary metathesis reactions. When an olefin of high reactivity is reacted with an olefin of lower reactivity (sterically bulky, electron-deficient, etc.), selective cross metathesis can be achieved using feedstock stoichiometries as low as 1:1. By employing a metathesis catalyst with the appropriate activity, selective cross metathesis reactions can be achieved with a wide variety of electron-rich, electron-deficient, and sterically bulky olefins. Application of this model has allowed for the prediction and development of selective cross metathesis reactions, culminating in unprecedented three-component intermolecular cross metathesis reactions.

Alkenes↗
PubMed1998

Bioanalytical calibration curves: proposal for statistical criteria.

Curve fitting procedures for bioanalytical assays are based on classical linear least squares (LSE) theory. A common procedure is to select among various models and weighting factors using the R2 as a goodness-of-fit criterion. It is questionable whether R2 is the most appropriate criterion for model selection. This is compounded by an often subjective removal of outliers. In this article, statistical curve fitting and diagnostic criteria are proposed. The fitting procedure is a Box-Cox-type power transformation of the data. The optimal transformation is obtained as the one that minimises the sum of squared deviations. Potential outlying standards are screened during the diagnostics stage as those whose jackknife percent deviations exceed 20%. The main advantage of this method is that it is objective and uniformly applicable across analytical techniques. Furthermore, the optimal transformation obtained in this way is unique. The results are demonstrated by comparing the power model to the R2 approach through the statistical analysis of 2094 analytical batches for 91 projects using various analytical techniques, namely GC, HPLC, LCMS and GCMS. The results indicate that the power model is robust and that QC batch acceptance using the power model is at least as good as the current method. These results hold true across all analytical techniques. It is thus strongly suggested that curve fitting and standard outlier detection for bioanalytical assays should be based on a power model and on jackknife percent deviations method with acceptable cut-off values.

Algorithms↗
PubMed2003

A local influence approach applied to binary data from a psychiatric study.

Recently, a lot of concern has been raised about assumptions needed in order to fit statistical models to incomplete multivariate and longitudinal data. In response, research efforts are being devoted to the development of tools that assess the sensitivity of such models to often strong but always, at least in part, unverifiable assumptions. Many efforts have been devoted to longitudinal data, primarily in the selection model context, although some researchers have expressed interest in the pattern-mixture setting as well. A promising tool, proposed by Verbeke et al. (2001, Biometrics 57, 43-50), is based on local influence (Cook, 1986, Journal of the Royal Statistical Society, Series B 48, 133-169). These authors considered the Diggle and Kenward (1994, Applied Statistics 43, 49-93) model, which is based on a selection model, integrating a linear mixed model for continuous outcomes with logistic regression for dropout. In this article, we show that a similar idea can be developed for multivariate and longitudinal binary data, subject to nonmonotone missingness. We focus on the model proposed by Baker, Rosenberger, and DerSimonian (1992, Statistics in Medicine 11, 643-657). The original model is first extended to allow for (possibly continuous) covariates, whereafter a local influence strategy is developed to support the model-building process. The model is able to deal with nonmonotone missingness but has some limitations as well, stemming from the conditional nature of the model parameters. Some analytical insight is provided into the behavior of the local influence graphs.

Antidepressive Agents, Tricyclic↗
PubMed2003

A model-fitting implementation of the DeFries-Fulker model for selected twin data.

In this research note, DeFries-Fulker (DF) regression analysis is reframed in model-fitting terms, where an individual's expected score is modeled as a function of their co-twin's proband status. This more flexible implementation of the DF model allows DZ-O twins to be incorporated in a sex-limitation model. Brief simulation results are presented along with the Mx scripts used.

Computer Simulation↗
PubMedDate not supplied

Effect of capsaicin and dimethyl sulfoxide on ion transport in the selected experimental models.

The aim of the present work was to determine the changes in ion transport in the selected epithelium-lined organs under influence of mechanical stimuli, and also to assess similarities and differences in reactions to capsaicin and dimethyl sulfoxide (DMSO) between trachea and caecum of rabbit and the skin of frog in this experimental setup. The experiments were conducted on rabbit trachea and caecum, and the skin of frog, Rana esculenta L. The experiments consisted in measuring transepithelial electrical potential (PD in mV) with Ussing apparatus, modified to enable testing of the effects of mechanical stimulation of organs and defined pharmacological treatments. It was demonstrated that the addition of DMSO to the stimulating fluid decreased reversible hyperpolarization (dPD) after mechanical stimulation by at least 50% in all studied groups. On the other hand, action of capsaicin was dependent on the organ studied as well as on experimental conditions (e.g. type of incubation). Capsaicin decreased PD and reaction to mechanical stimulation in trachea incubated in Ringer solution supplemented with amiloride. On the other hand, it did not influence electrophysiological parameters of the trachea following its incubation with bumetanide. Capsaicin did not change electrical potential or reactivity of rabbit caecum incubated with both amiloride and bumetanide. The administration of capsaicin on frog skin incubated with bumetanide caused inhibition of the reaction to mechanical stimulation, whereas during incubation with amiloride no changes were recorded in PD and dPD of the skin. The present study demonstrated that capsaicin and DMSO could modify processes of ion transport dependent on mechanical stimulation.

Animals↗
PubMed2005

Biased estimators of quantitative trait locus heritability and location in interval mapping.

In many empirical studies, it has been observed that genome scans yield biased estimates of heritability, as well as genetic effects. It is widely accepted that quantitative trait locus (QTL) mapping is a model selection procedure, and that the overestimation of genetic effects is the result of using the same data for model selection as estimation of parameters. There are two key steps in QTL modeling, each of which biases the estimation of genetic effects. First, test procedures are employed to select the regions of the genome for which there is significant evidence for the presence of QTL. Second, and most important for this demonstration, estimates of the genetic effects are reported only at the locations for which the evidence is maximal. We demonstrate that even when we know there is just one QTL present (ignoring the testing bias), and we use interval mapping to estimate its location and effect, the estimator of the effect will be biased. As evidence, we present results of simulations investigating the relative importance of the two sources of bias and the dependence of bias of heritability estimators on the true QTL heritability, sample size, and the length of the investigated part of the genome. Moreover, we present results of simulations demonstrating the skewness of the distribution of estimators of QTL locations and the resulting bias in estimation of location. We use computer simulations to investigate the dependence of this bias on the true QTL location, heritability, and the sample size.

Bias↗
PubMed2004

A two-state model for selective solubilization of benzene-limonene mixtures in sodium dihexyl sulfosuccinate microemulsions.

When surfactants are used to solubilize oil, the oil to be solubilized is often a mixture of components with differing properties, for example, solubilization of drug molecules in microemulsion formulations, remediation of organic polluted aquifers using surfactants, and so forth. Previous research has demonstrated that selective solubilization of one organic component over the other may occur if the organic components are dissimilar. In this research, we investigated selective solubilization from benzene-limonene mixtures in Winsor type I and III microemulsion systems containing water, sodium di-n-hexyl sulfosuccinate, and NaCl. The effect of the oil phase composition and the electrolyte concentration on the selectivity was studied. It was found that the selectivity toward benzene was highest at low electrolyte and benzene concentrations, decreasing as the electrolyte or benzene concentration increased. The results are discussed on the basis of the two-state solubilization theory and by correlating the curvature of the surfactant film in the microemulsion with changes of the electrolyte concentration and the oil phase composition. A simple mathematical model is developed for the selectivity, which combines the two-state solubilization theory and the net-average curvature model of microemulsion solubilization to yield close agreement with the experimental data.

Journal Article↗
PubMed2005

An index of local sensitivity to nonignorable drop-out in longitudinal modelling.

In longitudinal studies with potentially nonignorable drop-out, one can assess the likely effect of the nonignorability in a sensitivity analysis. Troxel et al. proposed a general index of sensitivity to nonignorability, or ISNI, to measure sensitivity of key inferences in a neighbourhood of the ignorable, missing at random (MAR) model. They derived detailed formulas for ISNI in the special case of the generalized linear model with a potentially missing univariate outcome. In this paper, we extend the method to longitudinal modelling. We use a multivariate normal model for the outcomes and a regression model for the drop-out process, allowing missingness probabilities to depend on an unobserved response. The computation is straightforward, and merely involves estimating a mixed-effects model and a selection model for the drop-out, together with some simple arithmetic calculations. We illustrate the method with three examples.

Animal Feed↗
PubMed2006

Robust Bayesian clustering.

A new variational Bayesian learning algorithm for Student-t mixture models is introduced. This algorithm leads to (i) robust density estimation, (ii) robust clustering and (iii) robust automatic model selection. Gaussian mixture models are learning machines which are based on a divide-and-conquer approach. They are commonly used for density estimation and clustering tasks, but are sensitive to outliers. The Student-t distribution has heavier tails than the Gaussian distribution and is therefore less sensitive to any departure of the empirical distribution from Gaussianity. As a consequence, the Student-t distribution is suitable for constructing robust mixture models. In this work, we formalize the Bayesian Student-t mixture model as a latent variable model in a different way from Svensén and Bishop [Svensén, M., & Bishop, C. M. (2005). Robust Bayesian mixture modelling. Neurocomputing, 64, 235-252]. The main difference resides in the fact that it is not necessary to assume a factorized approximation of the posterior distribution on the latent indicator variables and the latent scale variables in order to obtain a tractable solution. Not neglecting the correlations between these unobserved random variables leads to a Bayesian model having an increased robustness. Furthermore, it is expected that the lower bound on the log-evidence is tighter. Based on this bound, the model complexity, i.e. the number of components in the mixture, can be inferred with a higher confidence.

Algorithms↗
PubMed2002

Factors affecting the adsorption capacity of dissolved organic matter onto activated carbon: modified isotherm analysis.

Effects of aqueous phase Ca(II) concentrations and pH levels on the adsorption capacity of dissolved organic matters (DOMs) as well as the capacity dependency on the sources and types of DOMs and activated carbons (ACs) were examined. The Ca(II) effect was studied for three coal-based ACs having different pore size distributions (PSDs) and for three DOMs contained in three water samples. For each water sample, four working solutions prepared by adding different dosages of Ca(II) (0-1.5 mM) were used. For the water sample whose adsorption capacity was least affected by Ca(II), the effect of pH was subsequently examined for four water pH levels (pH = 5.5-10) and these three ACs. Isotherm data were analyzed in terms of a modified isotherm model selected from three models developed for normalizing isotherms of heterogeneous organic mixtures. The Ca(II) effect depended obviously on the DOMs used. For all three ACs, increasing Ca(II) greatly enhanced the adsorption capacity of a commercial humic acid. However, the capacity of the DOM in a naturally colored surface water source was much less affected and that of the DOM remaining after precoagulation was not affected. The DOM-Ca(II) interactions, rather than the AC-Ca(II) interactions, seemed to be the predominant mechanism that controlled the extent of the Ca(II)'s impact. Water pH greatly affected the adsorption of the surface water DOM remaining after precoagulation. This effect seemed to be attributed to pH's capability in changing ACs' charge characteristics reflected by zeta-potentials. The adsorption capacity increased as the water pH decreased and was well correlated as a function of pH and the volume of AC pores in sizes of 30-100 A. ACs having more pores in this size region generally exhibited larger adsorption capacities.

Adsorption↗
PubMedDate not supplied

Respiratory protection as a function of respirator fitting characteristics and fit-test accuracy.

The fitting characteristics of particulate respirators are no longer assessed in the National Institute for Occupational Safety and Health respirator certification program. It is important for respirator program administrators to understand the implications of that change and the additional burden it may impose. To address that issue, a typical respirator fit-testing program is analyzed using a mathematical model that describes the effectiveness of a fit-testing program as a function of the fitting characteristics of the respirator and the accuracy of the fittesting method. The model is used to estimate (1) the respirator assignment error, the percentage of respirator wearers mistakenly assigned an ill-fitting respirator; (2) the number of fit-test trials necessary to qualify a group of workers for respirator use; and (3) the number of workers who will fail the fit-test with any candidate respirator model and thereby fail to qualify for respirator use. Using data from previous studies, the model predicts respirator assignment errors ranging from 0 to 20%, depending on the fitting characteristics of the respirator models selected and the fit-testing method used. This analysis indicates that when respirators do not necessarily have good fitting characteristics, respirator program administrators should exercise increased care in the selection of respirator models and increased care in fit-testing. Also presented are ways to assess the fitting characteristics of candidate respirator models by monitoring the first-time fit-testing results. The model demonstrates that significant public health and economic benefits can result when only respirators having good fitting characteristics are purchased and respirators are assigned to workers using highly accurate fit-testing methods.

Dust↗