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 1,207 records · Page 67Linked to original sources
PubMed2006

Basal and hyperaemic myocardial blood flow in regionally denervated canine hearts: an in vivo study with positron emission tomography.

PURPOSE: Positron emission tomography (PET) studies in patients with diabetic autonomic neuropathy (DAN) have demonstrated the impact of this disease on cardiac sympathetic innervation and myocardial blood flow (MBF). To investigate the effects of selective partial sympathetic denervation of the left ventricle (LV) on baseline and hyperaemic MBF, we measured myocardial presynaptic catecholamine re-uptake (uptake-1), beta-adrenoceptor (beta-AR) density and MBF non-invasively by means of PET in a canine model of regional sympathetic denervation. METHODS: In 11 anaesthetised dogs, the sympathetic nerves of the free wall and septum of the LV were removed by means of dissection and phenol painting. Three weeks later, the animals were studied with PET. MBF was measured at baseline and following i.v. adenosine (140 microg kg(-1) min(-1)) and dobutamine (20 microg kg(-1) min(-1)) using(15)O-labelled water. Sympathetic denervation was confirmed by an 80+/-12% decrease in the volume of distribution (V(d)) of [(11)C]hydroxyephedrine (HED) compared with innervated regions. Myocardial beta-AR density was measured using [(11)C]CGP12177. RESULTS: Innervated and denervated regions showed no differences in MBF at baseline and during adenosine or dobutamine. [(11)C]HED V(d)was inversely correlated with MBF in both regions at baseline, and the correlation was lost during hyperaemia in denervated regions. However, for any given value of MBF, [(11)C]HED V(d)was significantly lower in the denervated regions. beta-AR density was comparable in denervated and innervated regions (17.9+/-4.2 vs 18.4+/-3.3 pmol g(-1); p=NS). CONCLUSION: In this experimental model, selective, regional sympathetic denervation of the LV, which results in a profound reduction in [(11)C]HED V(d), did not affect baseline or hyperaemic MBF. In addition, we demonstrated that, under baseline conditions, there was a significant inverse correlation between [(11)C]HED V(d)and MBF in both denervated and innervated regions.

Animals↗
PubMed2002

A Bayesian analysis of a proportion under non-ignorable non-response.

The National Health Interview Survey (NHIS) is one of the surveys used to assess one aspect of the health status of the U.S. population. One indicator of the nation's health is the total number of doctor visits made by the household members in the past year. We study the binary variable of at least one doctor visit versus no doctor visit by all household members to each of the 50 states and the District of Columbia. The proportion of households with at least one doctor visit is an indicator of the status of health of the U.S. population. There is a substantial number of non-respondents among the sampled households. The main issue we address here is that the non-response mechanism should not be ignored because respondents and non-respondents differ. The purpose of this work is to estimate the proportion of households with at least one doctor visit, and to investigate what adjustment needs to be made for non-ignorable non-response. We consider a non-ignorable non-response model that expresses uncertainty about ignorability through the ratio of odds of a household doctor visit among respondents to the odds of doctor visit among all households, and this ratio varies from state to state. We use a hierarchical Bayesian selection model to accommodate this non-response mechanism. Because of the weak identifiability of the parameters, it is necessary to 'borrow strength' across states as in small area estimation. We also perform a simulation study to compare the expansion model with an alternative expansion model, an ignorable model and a non-ignorable model. Inference for the probability of a doctor visit is generally similar across the models. Our main result is that for some of the states the non-response mechanism can be considered non-ignorable, and that 95 per cent credible intervals of the probability for a household doctor visit and the probability that a household responds shed important light on the NHIS data.

Bayes Theorem↗
PubMed2006

Exploration of methods to identify polymorphisms associated with variation in DNA repair capacity phenotypes.

Elucidating the relationship between polymorphic sequences and risk of common disease is a challenge. For example, although it is clear that variation in DNA repair genes is associated with familial cancer, aging and neurological disease, progress toward identifying polymorphisms associated with elevated risk of sporadic disease has been slow. This is partly due to the complexity of the genetic variation, the existence of large numbers of mostly low frequency variants and the contribution of many genes to variation in susceptibility. There has been limited development of methods to find associations between genotypes having many polymorphisms and pathway function or health outcome. We have explored several statistical methods for identifying polymorphisms associated with variation in DNA repair phenotypes. The model system used was 80 cell lines that had been resequenced to identify variation; 191 single nucleotide substitution polymorphisms (SNPs) are included, of which 172 are in 31 base excision repair pathway genes, 19 in 5 anti-oxidation genes, and DNA repair phenotypes based on single strand breaks measured by the alkaline Comet assay. Univariate analyses were of limited value in identifying SNPs associated with phenotype variation. Of the multivariable model selection methods tested: the easiest that provided reduced error of prediction of phenotype was simple counting of the variant alleles predicted to encode proteins with reduced activity, which led to a genotype including 52 SNPs; the best and most parsimonious model was achieved using a two-step analysis without regard to potential functional relevance: first SNPs were ranked by importance determined by random forests regression (RFR), followed by cross-validation in a second round of RFR modeling that included ever more SNPs in declining order of importance. With this approach six SNPs were found to minimize prediction error. The results should encourage research into utilization of multivariate analytical methods for epidemiological studies of the association of genetic variation in complex genotypes with risk of common diseases.

Cell Line↗
PubMed2006

Sexual selection and the evolution of evolvability.

Here we show that sexual selection can have an effect on the rate of mutation. We simulated the fate of a genetic modifier of the mutation rate in a sexual population with and without sexual selection (modelled using a female choice mechanism). Female choice for 'good genes' should reduce variability among male subjects, leaving insufficient differences to maintain female preferences. However, female choice can actually increase genetic variability by supporting a higher mutation rate in sexually selected traits. Increasing the mutation rate will be selected against because of the resulting decline in mean fitness. However, it also increases the probability of rare beneficial mutations arising, and mating skew caused by female preferences for male subjects carrying those beneficials with few deleterious mutations ('good genes') can lead to a mutation rate above that expected under natural selection. A choice of two male subjects was sufficient for there to be a twofold increase in the mutation rate as opposed to a decrease found under random mating.

Animals↗
PubMed2004

A multi-model approach to nucleic acid-based drug development.

With the advent of functional genomics and the shift of interest towards sequence-based therapeutics, the past decades have witnessed intense research efforts on nucleic acid-mediated gene regulation technologies. Today, RNA interference is emerging as a groundbreaking discovery, holding promise for development of genetic modulators of unprecedented potency. Twenty-five years after the discovery of antisense RNA and ribozymes, gene control therapeutics are still facing developmental difficulties, with only one US FDA-approved antisense drug currently available in the clinic. Limited predictability of target site selection models is recognized as one major stumbling block that is shared by all of the so-called complementary technologies, slowing the progress towards a commercial product. Currently employed in vitro systems for target site selection include RNAse H-based mapping, antisense oligonucleotide microarrays, and functional screening approaches using libraries of catalysts with randomized target-binding arms to identify optimal ribozyme/DNAzyme cleavage sites. Individually, each strategy has its drawbacks from a drug development perspective. Utilization of message-modulating sequences as therapeutic agents requires that their action on a given target transcript meets criteria of potency and selectivity in the natural physiological environment. In addition to sequence-dependent characteristics, other factors will influence annealing reactions and duplex stability, as well as nucleic acid-mediated catalysis. Parallel consideration of physiological selection systems thus appears essential for screening for nucleic acid compounds proposed for therapeutic applications. Cellular message-targeting studies face issues relating to efficient nucleic acid delivery and appropriate analysis of response. For reliability and simplicity, prokaryotic systems can provide a rapid and cost-effective means of studying message targeting under pseudo-cellular conditions, but such approaches also have limitations. To streamline nucleic acid drug discovery, we propose a multi-model strategy integrating high-throughput-adapted bacterial screening, followed by reporter-based and/or natural cellular models and potentially also in vitro assays for characterization of the most promising candidate sequences, before final in vivo testing.

Animals↗
PubMed2004

SJG-136 (NSC 694501), a novel rationally designed DNA minor groove interstrand cross-linking agent with potent and broad spectrum antitumor activity: part 2: efficacy evaluations.

Pyrrolo[2,1-c][1,4]benzodiazepine dimer SJG-136 (NSC 694501) selectively cross-links guanine residues located on opposite strands of DNA, and exhibits potent in vitro cytotoxicity. In addition, SJG-136 is highly active in vivo in hollow fiber assays. In the current investigation, SJG-136 was evaluated for in vivo efficacy in 10 tumor models selected on the basis of sensitivity of cells grown in the hollow fiber and in vitro time course assays: LOX IMVI and UACC-62 (melanomas); OVCAR-3 and OVCAR-5 (ovarian carcinomas); MDA-MB-435 (breast carcinoma); SF-295 and C-6 (gliomas); LS-174T (colon carcinoma); HL-60 TB (promyelocytic leukemia); and NCI-H522 (lung carcinoma). SJG-136 was active against small (150 mg) and large (250-400 mg) xenografts with tumor mass reductions in all 10 models. In addition, significant growth delays occurred in nine models, cell kill in six models ranged between 1.9 and 7.2 logs, and there were 1 to 4/6 tumor-free responses in six models. SJG-136 is active following i.v. bolus injections, as well as by 5-day continuous infusions. Of all of the schedules tested, bolus administrations for 5 consecutive days (qd x 5) conferred the greatest efficacy. SJG-136 is active over a wide dosage range in athymic mouse xenografts: on a qd x 5 schedule, the maximum-tolerated dose was approximately 120 microg/kg/dose (total dose: 0.6 mg/kg = 1.8 mg/m2) and the minimum effective dose in the most sensitive model (SF-295) was approximately 16 microg/kg/dose (total dose: 0.08 mg/kg = 0.24 mg/m2). Results of this study extend the initial in vivo observations reported in the reference above and confirm the importance of expediting more detailed preclinical evaluations on this novel agent in support of phase I clinical trials in the United Kingdom and the United States, which are planned to commence shortly.

Animals↗
PubMed2004

Bayesian class discovery in microarray datasets.

A novel approach to class discovery in gene expression datasets is presented. In the context of clinical diagnosis, the central goal of class discovery algorithms is to simultaneously find putative (sub-)types of diseases and to identify informative subsets of genes with disease-type specific expression profile. Contrary to many other approaches in the literature, the method presented implements a wrapper strategy for feature selection, in the sense that the features are directly selected by optimizing the discriminative power of the used partitioning algorithm. The usual combinatorial problems associated with wrapper approaches are overcome by a Bayesian inference mechanism. On the technical side, we present an efficient optimization algorithm with guaranteed local convergence property. The only free parameter of the optimization method is selected by a resampling-based stability analysis. Experiments with Leukemia and Lymphoma datasets demonstrate that our method is able to correctly infer partitions and corresponding subsets of genes which both are relevant in a biological sense. Moreover, the frequently observed problem of ambiguities caused by different but equally high-scoring partitions is successfully overcome by the model selection method proposed.

Algorithms↗
PubMed1996

Maximum likelihood analysis of rare binary traits under different modes of inheritance.

Maximum likelihood methodology was applied to determine the mode of inheritance of rare binary traits with data structures typical for swine populations. The genetic models considered included a monogenic, a digenic, a polygenic, and three mixed polygenic and major gene models. The main emphasis was on the detection of major genes acting on a polygenic background. Deterministic algorithms were employed to integrate and maximize likelihoods. A simulation study was conducted to evaluate model selection and parameter estimation. Three designs were simulated that differed in the number of sires/number of dams within sires (10/10, 30/30, 100/30). Major gene effects of at least one SD of the liability were detected with satisfactory power under the mixed model of inheritance, except for the smallest design. Parameter estimates were empirically unbiased with acceptable standard errors, except for the smallest design, and allowed to distinguish clearly between the genetic models. Distributions of the likelihood ratio statistic were evaluated empirically, because asymptotic theory did not hold. For each simulation model, the Average Information Criterion was computed for all models of analysis. The model with the smallest value was chosen as the best model and was equal to the true model in almost every case studied.

Algorithms↗
PubMedDate not supplied

Three-dimensional model of a selective theophylline-binding RNA molecule.

A three-dimensional (3D) model for an RNA molecule that selectively binds theophylline but not caffeine is proposed. This RNA, which was found using SELEX (Jenison et al., 1994), is 10,000 times more specific for theophylline (Kn = 320 nM) than for caffeine (KD = 3.5 mM), although the two ligands are identical except for a methyl group substituted at N7 (present only in caffeine). The binding affinity for ten xanthine-based ligands was used to derive a comparative molecular field analysis model (R2 = 0.93 for three components, with cross-validated R2 of 0.73), using the SYBYL and GOLPE programs. A pharmacophoric map was generated to locate steric and electrostatic interactions between theophylline and the RNA binding site. This information was used to identify putative functional groups of the binding pocket and to generate distance constraints. On the basis of a model for the secondary structure (Jenison et al., 1994), the 3D structure of this RNA was then generated using the following method: each helical region of the RNA molecule was treated as a rigid body; single-stranded loops with specific end-to-end distances were generated. The structures of RNA-xanthine complexes were studied using a modified Monte Carlo algorithm. The detailed structure of an RNA-ligand complex model, as well as possible explanations for the theophylline selectivity are discussed.

Base Sequence↗
PubMed2002

Evidence that selected amplification of a bacterial lac frameshift allele stimulates Lac(+) reversion (adaptive mutation) with or without general hypermutability.

In the genetic system of Cairns and Foster, a nongrowing population of an E. coli lac frameshift mutant appears to specifically accumulate Lac(+) revertants when starved on medium including lactose (adaptive mutation). This behavior has been attributed to stress-induced general mutagenesis in a subpopulation of starved cells (the hypermutable state model). We have suggested that, on the contrary, stress has no direct effect on mutability but favors only growth of cells that amplify their leaky mutant lac region (the amplification mutagenesis model). Selection enhances reversion primarily by increasing the mutant lac copy number within each developing clone on the selection plate. The observed general mutagenesis is attributed to a side effect of growth with an amplification-induction of SOS by DNA fragments released from a tandem array of lac copies. Here we show that the S. enterica version of the Cairns system shows SOS-dependent general mutagenesis and behaves in every way like the original E. coli system. In both systems, lac revertants are mutagenized during selection. Eliminating the 35-fold increase in mutation rate reduces revertant number only 2- to 4-fold. This discrepancy is due to continued growth of amplification cells until some clones manage to revert without mutagenesis solely by increasing their lac copy number. Reversion in the absence of mutagenesis is still dependent on RecA function, as expected if it depends on lac amplification (a recombination-dependent process). These observations support the amplification mutagenesis model.

Bacterial Proteins↗
PubMed1993

A novel factor Xa inhibitor: structure-activity relationships and selectivity between factor Xa and thrombin.

A series of 3-amidinoaryl-2-[4-[ [(3S)-3-pyrrolidinyl]oxy]phenyl] propanoic acids have been investigated for development of a novel factor Xa inhibitor, possessing a potent inhibitory activity for factor Xa and a selectivity for factor Xa compared to thrombin. In order to study the structure-activity relationships and the selectivity, models of factors Xa complexes formed with the inhibitors were constructed on the basis of X-ray crystallographic data of a trypsin-inhibitor complex. The models showed that the binding mode of the inhibitors to the S1 pocket of the enzyme accounted for the structure-activity relationships and that the difference between Gln192 of factor Xa and Glu192 of thrombin had a key role in the selectivity.

Amidines↗
PubMed2003

A Bayesian neural network approach for modelling censored data with an application to prognosis after surgery for breast cancer.

A Bayesian framework is introduced to carry out Automatic Relevance Determination (ARD) in feedforward neural networks to model censored data. A procedure to identify and interpret the prognostic group allocation is also described. These methodologies are applied to 1616 records routinely collected at Christie Hospital, in a monthly cohort study with 5-year follow-up. Two cohort studies are presented, for low- and high-risk patients allocated by standard clinical staging. The results of contrasting the Partial Logistic Artificial Neural Network (PLANN)-ARD model with the proportional hazards model are that the two are consistent, but the neural network may be more specific in the allocation of patients into prognostic groups. With automatic model selection, the regularised neural network is more conservative than the default stepwise forward selection procedure implemented by SPSS with the Akaike Information Criterion.

Bayes Theorem↗
PubMed1993

Purine nucleoside phosphorylase. Catalytic mechanism and transition-state analysis of the arsenolysis reaction.

Purine nucleoside phosphorylase from calf spleen catalyzes the arsenolysis of inosine to form hypoxanthine and ribose 1-arsenate, which spontaneously hydrolyzes to ribose and arsenate. In the presence of H2(18)O, no 18O is incorporated into ribose, demonstrating that ribose 1-arsenate hydrolysis occurs by attack of water on the arsenic atom. Rapid reaction kinetics at 20 degrees C result in a biphasic rate curve with the first turnover occurring at a rate of 20 s-1 followed by a steady-state rate of 2 s-1. The product burst is consistent with rapid steps for substrate binding and arsenolysis followed by rate-limiting hypoxanthine release at a rate of 2 s-1. Purine nucleoside phosphorylase with bound [14C]inosine was mixed with excess unlabeled inosine and arsenate to determine relative rates for reaction or dissociation of bound inosine. The commitment factor (product formed/inosine released) was 0.19 at saturating arsenate, indicating that inosine binds to free enzyme and that bound inosine is not in thermodynamic equilibrium with free substrate. At neutral pH, kinetic isotope effects for the phosphorolysis reaction are small, indicating kinetic suppression. Kinetic isotope effects for arsenolysis were measured with [1'-3H]-, [2'-3H]-, [1'-14C]-, [9-15N]-, [4'-3H]-, and [5'-3H]inosine to provide experimental values of 1.118 +/- 0.003, 1.128 +/- 0.003, 1.022 +/- 0.005, 1.009 +/- 0.004, 1.007 +/- 0.003 and 1.028 +/- 0.004 respectively. Following correction for commitment factors, the intrinsic isotope effects were matched to a geometric transition-state model selected by bond-energy bond order vibrational analysis.(ABSTRACT TRUNCATED AT 250 WORDS)

Animals↗
PubMed2006

A better lemon squeezer? Maximum-likelihood regression with beta-distributed dependent variables.

Uncorrectable skew and heteroscedasticity are among the "lemons" of psychological data, yet many important variables naturally exhibit these properties. For scales with a lower and upper bound, a suitable candidate for models is the beta distribution, which is very flexible and models skew quite well. The authors present maximum-likelihood regression models assuming that the dependent variable is conditionally beta distributed rather than Gaussian. The approach models both means (location) and variances (dispersion) with their own distinct sets of predictors (continuous and/or categorical), thereby modeling heteroscedasticity. The location sub-model link function is the logit and thereby analogous to logistic regression, whereas the dispersion sub-model is log linear. Real examples show that these models handle the independent observations case readily. The article discusses comparisons between beta regression and alternative techniques, model selection and interpretation, practical estimation, and software.

Analysis of Variance↗
PubMed2002

Bayesian hierarchical models for multi-level repeated ordinal data using WinBUGS.

Multi-level repeated ordinal data arise if ordinal outcomes are measured repeatedly in subclusters of a cluster or on subunits of an experimental unit. If both the regression coefficients and the correlation parameters are of interest, the Bayesian hierarchical models have proved to be a powerful tool for analysis with computation being performed by Markov Chain Monte Carlo (MCMC) methods. The hierarchical models extend the random effects models by including a (usually flat) prior on the regression coefficients and parameters in the distribution of the random effects. Because the MCMC can be implemented by the widely available BUGS or WinBUGS software packages, the computation burden of MCMC has been alleviated. However, thoughtfulness is essential in order to use this software effectively to analyze such data with complex structures. For example, we may have to reparameterize the model and standardize the covariates to accelerate the convergence of the MCMC, and then carefully monitor the convergence of the Markov chain. This article aims at resolving these issues in the application of the WinBUGS through the analysis of a real multi-level ordinal data. In addition, we extend the hierarchical model to include a wider class of distributions for the random effects. We propose to use the deviance information criterion (DIC) for model selection. We show that the WinBUGS software can readily implement such extensions and the DIC criterion.

Bayes Theorem↗
PubMed2005

TM-align: a protein structure alignment algorithm based on the TM-score.

We have developed TM-align, a new algorithm to identify the best structural alignment between protein pairs that combines the TM-score rotation matrix and Dynamic Programming (DP). The algorithm is approximately 4 times faster than CE and 20 times faster than DALI and SAL. On average, the resulting structure alignments have higher accuracy and coverage than those provided by these most often-used methods. TM-align is applied to an all-against-all structure comparison of 10 515 representative protein chains from the Protein Data Bank (PDB) with a sequence identity cutoff <95%: 1996 distinct folds are found when a TM-score threshold of 0.5 is used. We also use TM-align to match the models predicted by TASSER for solved non-homologous proteins in PDB. For both folded and misfolded models, TM-align can almost always find close structural analogs, with an average root mean square deviation, RMSD, of 3 A and 87% alignment coverage. Nevertheless, there exists a significant correlation between the correctness of the predicted structure and the structural similarity of the model to the other proteins in the PDB. This correlation could be used to assist in model selection in blind protein structure predictions. The TM-align program is freely downloadable at http://bioinformatics.buffalo.edu/TM-align.

Algorithms↗
PubMed1997

Genetic selection strategies: computer modeling.

There are four primary factors to consider in genetic selection strategies: 1) accuracy of selection, 2) selection intensity, 3) effective population size, and 4) mating system. Current theory indicates that optimum response to selection is achieved by maximizing the first three factors and using a mating systems that allows optimization of reproductive characteristics in dam lines and production characteristics in sire lines. However, with limited resources, compromises among the first three factors are needed. Simulations are useful for examining those compromises. Unrealistic simplifying assumptions are necessary for analytic theoretical results and thus do not address real world breeding problems. Using simulations, the relationship between selection accuracy, which is increased by use of family selection indices or Best Linear Unbiased Prediction (BLUP), and response to selection was examined. Results show that those procedures place a great restriction on effective population size, which offsets most of their advantage, i.e., there is too little emphasis on effective population size. A revision of the methodology and a reappraisal of the results of selection theory for optimization of genetic response is required. Another relationship that is of fundamental importance in breeding programs is that between selection intensity and effective population size. Analytical results for the additive case have been developed but have never been extended to heterotic traits. A gene level simulation program was developed to examine that relationship. Results show that the optimal selection strategy depends on the trait being selected. For additive traits and in the short term (20 generations), one should maximize selection intensity. For heterotic traits, an intermediate proportion (25% of each sex) gives optimal response. In all breeding strategies, primary attention must be given to the rate of inbreeding, which is increased by increasing either accuracy of selection or selection intensity. Inbreeding reduces response to selection in two ways. First, for both additive and nonadditive traits, inbreeding is a measure of the amount of random genetic drift that has occurred. Genetic drift causes loss of favorable alleles. Once lost, those alleles can never be recovered and thus genetic drift lowers the selection limit. Second, for heterotic traits, inbreeding results in a depression of the mean caused by directional dominance.

Animals↗
PubMed1994

Superinfection and the evolution of parasite virulence.

Earlier ideas that parasites evolve toward becoming harmless to their hosts have, in recent years, given way to more analytic studies, focused on the 'basic reproductive rate', R0, of individual parasites. In general, the biology of the parasite life cycle will lead to constraining relations between virulence (parasite-associated host death or reduction in fertility) and transmissibility: the maximum R0 may then be attained by virulence being high, or low, or at some intermediate level, depending on the details of the constraining relations. Such studies have not generally included superinfection (where an already-infected host is infected by another parasite). Here we propose a general, but simple, model of superinfection, which is amenable to analytical treatment. In such models selection does not simply act to maximize R0; superinfection leads to selection for higher levels of virulence, highly polymorphic parasite populations and very complicated dynamics. We calculate the equilibrium distribution of parasite strains and the maximum level of virulence that can be maintained by superinfection. We also note the equivalence between our 'superinfection model' and recent approaches to the study of the meta-population dynamics of multi-species interactions.

Animals↗