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 1,153 records · Page 64Linked to original sources
PubMed2003

Comparison of blood particle deposition models for non-parallel flow domains.

Adhesions of monocytes and platelets to a vascular surface, particularly in regions of flow stagnation, recirculation, and reattachment, are a significant initial event in a broad spectrum of particle-wall interactions that significantly influence the formation of stenotic lesions and mural thrombi. A number of approximations are available for the simulation of both monocyte and platelet interactions with the vascular surface. For the simulation of blood particle adhesion, this study hypothesizes that: (a) the discrete element approach, which accounts for finite particle size and inertia, is advantageous in the context of non-parallel flow domains including stagnation, recirculation, and reattachment; and (b) the likelihood for particle deposition may be effectively approximated as being non-linearly proportional to local particle concentration, residence time, and wall proximity. Models such as wall shear stress correlations, the multicomponent mixture approach, and Lagrangian particle tracking with and without hydrodynamic particle-wall interactions were evaluated. Quantitative performance of the selected models was established by comparisons to available experimental data sets for non-parallel axisymmetric suspension flows of monocytes and platelets. Factors including the convective-diffusive transport of particles, finite particle size and inertia, as well as near-wall hydrodynamic interactions were found to significantly influence blood particle deposition. Of the models studied, the near-wall residence time approach was found to be a particularly effective indicator for the deposition of monocytes (r2=0.74) and platelets (r2=0.57), given that nano-scale physical and biochemical effects must be greatly approximated in computational simulations involving relatively large-scale geometries and complex flow fields.

Animals↗
PubMed2005

In vitro cytotoxicity testing of airborne formaldehyde collected in serum-free culture media.

The purpose of this study was to identify a suitable sampling model for on-site toxicity assessment of soluble air contaminants such as formaldehyde, a well known industrial and indoor air contaminant. The in vitro cytotoxicity of formaldehyde, the selected model for soluble air contaminants, was studied using the MTS (tetrazolium salt) assay in two carcinoma cell lines, A549 epithelial lung and HepG2 hepatocarcinoma, and in skin fibroblasts. The cytotoxic effects of airborne formaldehyde were evaluated using test atmospheres in concentrations below 10 ppm (12.3 mg/m3), generated by a dynamic diffusion method and bubbled (0.3 L/min) through serum-free culture media for one or four hours. Human cells were treated with formaldehyde air samples, and cell viability was determined after four hours incubation. In parallel, the concentration of airborne formaldehyde was monitored, using the 3500 NIOSH method. Cell viability of the HepG2 cells exposed to formaldehyde air samples (8.75 ppm x 4 h) was reduced to less than 50% (31.6 +/- 1.24%). The HepG2 cell lines were found to be more sensitive (IC50 = 103.79 +/- 23.55 mg/L) to formaldehyde than both A549 cell lines (IC50= 198.36 +/- 9.54 mg/L) and skin fibroblasts (IC50 = 196.68 +/- 36.73 mg/L) (P < 0.01). An average of 96.8% was determined for collection efficiency of formaldehyde in serum-free culture media. The results of this study suggest that absorption of soluble air contaminants, such as formaldehyde, in serum-free culture media can be used as a suitable sampling model for on-site toxicity assessments.

Air Pollutants, Occupational↗
PubMed2003

Joint regression and association modeling of longitudinal ordinal data.

We propose models for longitudinal, or otherwise clustered, ordinal data. The association between subunit responses is characterized by dependence ratios (Ekholm, Smith, and McDonald, 1995, Biometrika 82, 847-854), which are extended from the binary to the multicategory case. The joint probabilities of the subunit responses are expressed as explicit functions of the marginal means and the dependence ratios of all orders, obtaining a computational advantage for likelihood-based inference. Equal emphasis is put on finding regression models for the univariate cumulative probabilities, and on deriving the dependence ratios from meaningful association-generating mechanisms. A data set on the effects of treatment with Fluvoxamine, which has been analyzed in parts before (Molenberghs, Kenward, and Lesaffre, 1997, Biometrika 84, 33-44), is analyzed in its entirety. Selection models are used for studying the sensitivity of the results to drop-out.

Analysis of Variance↗
PubMed2007

A residuals-based transition model for longitudinal analysis with estimation in the presence of missing data.

We propose a transition model for analysing data from complex longitudinal studies. Because missing values are practically unavoidable in large longitudinal studies, we also present a two-stage imputation method for handling general patterns of missing values on both the outcome and the covariates by combining multiple imputation with stochastic regression imputation. Our model is a time-varying auto-regression on the past innovations (residuals), and it can be used in cases where general dynamics must be taken into account, and where the model selection is important. The entire estimation process was carried out using available procedures in statistical packages such as SAS and S-PLUS. To illustrate the viability of the proposed model and the two-stage imputation method, we analyse data collected in an epidemiological study that focused on various factors relating to childhood growth. Finally, we present a simulation study to investigate the behaviour of our two-stage imputation procedure.

Bias↗
PubMed2005

Should phylogenetic models be trying to "fit an elephant"?

For the past two decades, there has been an ongoing debate within the plylogenetics community over whether model-based approaches for molecular systematics (such as maximum likelihood) should be preferred over the more traditional "maximum parsimony" approach. A recent simulation study by Kolaczkowski and Thornton has brought this debate into sharp focus. In this article, I discuss the significance of their findings and offer a prognosis on the implications for molecular phylogenetics. I believe that biochemistry and model selection have an important role in developing accurate phylogenetic approaches.

Animals↗
PubMed2005

Mixtures-of-experts of autoregressive time series: asymptotic normality and model specification.

We consider a class of nonlinear models based on mixtures of local autoregressive time series. At any given time point, we have a certain number of linear models, denoted as experts, where the vector of covariates may include lags of the dependent variable. Additionally, we assume the existence of a latent multinomial variable, whose distribution depends on the same covariates as the experts, that determines which linear process is observed. This structure, denoted as mixture-of-experts (ME), is considerably flexible in modeling the conditional mean function, as shown by Jiang and Tanner. In this paper, we present a formal treatment of conditions to guarantee the asymptotic normality of the maximum likelihood estimator (MLE), under stationarity and nonstationarity, and under correct model specification and model misspecification. The performance of common model selection criteria in selecting the number of experts is explored via Monte Carlo simulations. Finally, we present applications to simulated and real data sets, to illustrate the ability of the proposed structure to model not only the conditional mean, but also the whole conditional density.

Algorithms↗
PubMed1999

A comparison of three polytomous item response theory models in the context of testlet scoring.

An alternative to dichotomous scoring of multiple items anchored to a common stem is scoring these items as a single polytomous item (testlet scoring). This study systematically compared the partial credit model (PCM), the generalized partial credit model (GPCM), and the graded response model (GRM) in the context of testlet scoring. Data sets included a sample from the fall 1994 administration of the SAT I (N = 2,548) and a simulated data set. Theta estimation, information, and model fit were analyzed. Correlations among theta estimates ranged from 0.9748 to 0.9921. The relationship among the information functions of the PCM, GPCM and the GRM reflected the discrimination parameter estimates for the latter two models. Suggestions are made with regard to model selection.

Aptitude Tests↗
PubMed1997

Maintenance of transposable element copy number in natural populations of Drosophila melanogaster and D. simulans.

To investigate the main forces controlling the containment of transposable elements (TE) in natural populations, we analyzed the copia, mdg1, and 412 elements in various populations of Drosophila melanogaster and D. simulans. A lower proportion of insertion sites on the X chromosome in comparison with the autosomes suggests that selection against the detrimental effects of TE insertions is the major force containing TE copies in populations of Drosophila. This selection effect hypothesis is strengthened by the absence of the negative correlation between recombinaiton rate and TE copy number along the chromosomes, which was expected under the alternative ectopic exchange model (selection against the deleterious rearrangements promoted by recombination between TE insertions). A cline in 412 copy number in relation to latitude was observed among the natural populations of D. simulans, with very high numbers existing in some local populations (around 60 copies in a sample from Canberra, Australia). An apparent absence of selection effects in this Canberra sample and a value of transposition rate equal to 1-2 x 10(-3) whatever the population and its copy number agree with the idea of recent but temporarily drastic TE movements in local populations. The high values of transposition rate in D. simulans clearly disfavor the hypothesis that the low amount of transposable elements in this species could result from a low transposition rate.

Animals↗
PubMed1993

Challenges to dairy cattle management: genetic considerations.

Several new technologies will be used to produce continued genetic change in dairy cattle. These technologies are categorized broadly as 1) improved modeling, selection, and evaluation methods; 2) use of new and improved reproductive technologies; 3) new developments in molecular genetics; and 4) new developments in immunogenetics. Improvements in evaluation will continue as computers become faster and have more storage capabilities. Improved mathematical models that more nearly describe the biology of lactation will maximize estimation of genetic differences and reduce residuals. New reproductive technology could allow reduction of generation intervals two- to fivefold compared with present generation intervals and, combined with genetic markers, could markedly accelerate progress. Health problems in dairy cattle are expected to increase as production increases. Thus, selection for decreased incidences of health disorders will be needed, probably by selection of sires with improved general immunocompetence. Research is in the early stages of application of techniques of molecular genetics to animal breeding. Early uses will allow detection and alleviation of genetic defects. Eventually, marker genes that directly affect production and metabolic pathways that also affect production will be subjected to selection. The ability to foresee new and potentially useful techniques will be determined by scientific advancement of areas in which researchers are engaged; thus, accurate prediction far into the future cannot be expected.

Animals↗
PubMed1994

Chimeric evolution of the 2-microns genome in Saccharomyces cerevisiae.

We compared the nucleotide substitution pattern over the entire genome of two unique variants of the 6,300-bp selfish DNA (2 microns) plasmid in Saccharomyces cerevisiae. The DNA sequence of the left-unique region is identical among 2-microns variants, while the right-unique region shows substantial divergence. This chimeric pattern cannot be explained by neutral or Darwinian selection models. We propose that horizontal transmission of the 2-microns plasmid coupled with a directed, polarized gene conversion maintains the DNA sequence of the left-unique region, whereas the right-unique region is subject to random drift and Darwinian selection.

Base Sequence↗
PubMed2006

Seasonal confounding in air pollution and health time-series studies: effect on air pollution effect estimates.

A major statistical challenge in air pollution and health time-series studies is to adequately control for confounding effects of time-varying covariates. Daily health outcome counts are most commonly analysed by Poisson regression models, adjusted for overdispersion, with air pollution levels included as a linear predictor and smooth functions for calendar time and weather variables to adjust for time-varying confounders. Various smoothers have been used so far, but the optimal strategy for choosing smoothers and their degree of smoothing remains controversial. In this work, we evaluate the performance of various smoothers with different criteria for choosing the degree of smoothing in terms of bias and efficiency of the air pollution effect estimate in a simulation study. The evaluated approaches were also applied to real mortality data from 22 European cities. The simulation study imitated a multi-city study. Data were generated from a fully parametric model. Model selection methods which optimize prediction may lead to increased biases in the air pollution effect estimate. Minimization of the absolute value of the sum of the partial autocorrelation function of the model's residuals (PACF), as a criterion to choose the degree of smoothness, gave the smallest biases. The penalized splines (PS) method with a large number of effective dfs (e.g. 8-12 per year) could be used as the basic, relatively conservative, analysis whereas the PS and natural splines in combination with PACF could be applied to provide a reasonable range of the effect estimate.

Air Pollution↗
PubMed2026

Decoding glioblastoma evolution and heterogeneity through mechanistic modeling: implications for clinical translation.

Glioblastoma (GBM) is one of the most aggressive and lethal primary brain tumors in adults, characterized by dynamic clonal evolution and extensive genomic, cellular, spatial, and microenvironmental heterogeneity. Multi-omics studies have revealed that GBM follows complex evolutionary trajectories involving genetic, epigenetic, transcriptional, and immune-microenvironmental remodeling as tumors grow, adapt to the brain microenvironment, and acquire therapeutic resistance. Increasing evidence suggests that GBM may originate from aberrant neural stem or progenitor cells, including those residing in the subventricular zone, and that glioblastoma stem cells (GSCs) contribute to tumor propagation, heterogeneity, and recurrence. A key conceptual challenge is to reconcile hierarchical cancer stem cell models, in which GSCs are viewed as relatively stable tumor-propagating subpopulations, with dynamic state plasticity models, in which stem-like properties can be reversibly acquired or lost during transitions among proneural-like, mesenchymal-like, invasive, and therapy-tolerant states. Recent advances in single-cell profiling, spatial transcriptomics, lineage tracing, organoid culture, 3D bioprinting, genetically engineered models, and artificial intelligence (AI)-assisted computational modeling have substantially improved the ability to study these processes. However, no currently available model fully recapitulates human GBM heterogeneity, recurrence, treatment history, and tumor-microenvironment interactions. Therefore, model selection should be guided by clearly defined mechanistic questions rather than by reliance on any single platform. This review summarizes current advances in in vitro, ex vivo, in vivo, and computational models for studying GBM evolution and heterogeneity, and discusses how integrated model pipelines may improve preclinical drug testing, treatment-response prediction, and precision neuro-oncology.

Humans↗
PubMed2005

Modeling individual differences in cognition.

Many evaluations of cognitive models rely on data that have been averaged or aggregated across all experimental subjects, and so fail to consider the possibility of important individual differences between subjects. Other evaluations are done at the single-subject level, and so fail to benefit from the reduction of noise that data averaging or aggregation potentially provides. To overcome these weaknesses, we have developed a general approach to modeling individual differences using families of cognitive models in which different groups of subjects are identified as having different psychological behavior. Separate models with separate parameterizations are applied to each group of subjects, and Bayesian model selection is used to determine the appropriate number of groups. We evaluate this individual differences approach in a simulation study and show that it is superior in terms of the key modeling goals of prediction and understanding. We also provide two practical demonstrations of the approach, one using the ALCOVE model of category learning with data from four previously analyzed category learning experiments, the other using multidimensional scaling representational models with previously analyzed similarity data for colors. In both demonstrations, meaningful individual differences are found and the psychological models are able to account for this variation through interpretable differences in parameterization. The results highlight the potential of extending cognitive models to consider individual differences.

Cognition↗
PubMed2005

A statistical framework for haplotype block inference.

The existence of haplotype blocks transmitted from parents to offspring has been suggested recently. This has created an interest in the inference of the block structure and length. The motivation is that haplotype blocks that are characterized well will make it relatively easier to quickly map all the genes carrying human diseases. To study the inference of haplotype block systematically, we propose a statistical framework. In this framework, the optimal haplotype block partitioning is formulated as the problem of statistical model selection; missing data can be handled in a standard statistical way; population strata can be implemented; block structure inference/hypothesis testing can be performed; prior knowledge, if present, can be incorporated to perform a Bayesian inference. The algorithm is linear in the number of loci, instead of NP-hard for many such algorithms. We illustrate the applications of our method to both simulated and real data sets.

Algorithms↗
PubMed2005

[Preliminary study on biological behavior and staging of transplanted hepatoma in rats].

OBJECTIVE: To establish a staging standard for a transplanted hepatoma model in rats, which may be equivalent to the human primary liver cancer. METHODS: A transplanted hepatoma model was made in SD rats by implantation of histologically intact Walker-256 tumor fragment into the left lateral lobe of the liver. The body weight, tumor volume and serum albumin level of the rats were measured and assayed in different periods. Pathological sections were also fabricated to detect the invasion and metastasis of the neighboring tissues. RESULTS: The body weight loss and ascites occurred at the 12th and 18th day separately. The serum albumin level was (35.32+/-2.45) g/L at the 11th day, which showed no statistical difference as compared with the healthy rats (P>0.05), and the albumin level at the 18th day was significantly lower than that at the 11th day (P<0.05). At the 23rd day, 56.25% of the rats developed bloody ascites. CONCLUSION: The period between 5-11 days in the model rats may be correspondent to the early stage of human primary liver cancer, and the biological behavior of the model rats at the18th day may be correspondent to the late stage of human primary liver cancer. This staging standard might be the guideline for model selection in experimental study of primary liver cancer.

Animals↗
PubMed2006

Fat-tailed gene flow in the dioecious canopy tree species Fraxinus mandshurica var. japonica revealed by microsatellites.

Pollen flow, seed dispersal and individual reproductive success can be simultaneously estimated from the genotypes of adults and offspring using stochastic models. Using four polymorphic microsatellite loci, gene flow of the wind-pollinated and wind-seed-dispersed dioecious tree species, Fraxinus mandshurica var. japonica, was quantified in a riparian forest, in northern Japan. In a 10.5-ha plot, 74 female adults, 76 male adults and 292 current-year seedlings were mapped and genotyped, together with 200 seeds. To estimate dispersal kernels of pollen and seeds, we applied normal, exponential power, Weibull, bivariate t-distribution kernels, and two-component models consisting of two normal distribution functions, one with a small and one with a large variance. A two-component pollen flow model with a small contribution (26.1%) from short-distance dispersal (sigma = 7.2 m), and the rest from long-distance flow (sigma = 209.9 m), was chosen for the best-fitting model. The average distance that integrated pollen flows inside and outside the study plot was estimated to be 196.8 m. Tree size and flowering intensity affected reproduction, and there appeared to be critical values that distinguished reproductively successful and unsuccessful adults. In contrast, the gene flow model that estimated both pollen and seed dispersal from established seedlings resulted in extensive seed dispersal, and the expected spatial genetic structures did not satisfactorily fit with the observations, even for the selected model. Our results advanced small-scale individual-based parentage analysis for quantifying fat-tailed gene flow in wind-mediated species, but also clarified its limitations and suggested future possibilities for gene flow studies.

Alleles↗
PubMed1993

Pharmacodynamic-pharmacokinetic relationships and therapeutic drug monitoring.

Pharmacokinetic-pharmacodynamic studies are becoming increasingly important in the development of new anti-cancer drugs. The Hill maximal effect model describes a sigmoidal dose-response relationship and has been applied to analyses of both haematological and non-haematological toxicity. This review discusses several approaches to population pharmacodynamics, including the two stage, NONMEM, and non-parametric approaches. Pharmacodynamic models for the haematological toxicity of amonafide, carboplatin, doxorubicin, etoposide, HMBA and menogaril are discussed, as are models for non-haematological toxicity. Adaptive control methods and therapeutic drug monitoring are useful in dosing drugs with narrow therapeutic windows, but the indications for using such strategies should be carefully selected. Models for 5FU, HMBA, methotrexate, 6-mercaptopurine, carboplatin and etoposide are discussed. Limited sampling strategies can facilitate the completion of pharmacokinetic studies and should be developed during phase I testing of new compounds. A new area of future importance is the investigation of drugs with active metabolites, such as the anthracyclines and amonafide.

Antineoplastic Agents↗
PubMed2001

Human hepatocyte culture systems for the in vitro evaluation of cytochrome P450 expression and regulation.

Primary cultures of human hepatocytes have been used extensively by both academic and industrial laboratories for evaluating the hepatic disposition of drugs and other xenobiotics. Their primary utility has been for assessing the induction potential of new chemical entities (NCEs) and they continue to serve as the gold standard. Primary considerations for conducting in vitro drug testing utilizing cultures of human hepatocytes, such as the effects of culture and study conditions, are discussed. The maintenance of normal cellular physiology and intercellular contacts in vitro is of particular importance for optimal phenotypic gene expression and response to drugs and other xenobiotics. Significant advances in our understanding of cytochrome P450 (CYP450) enzyme regulation have been made with the recent identification of the nuclear receptors mediating the induction of CYP2B and CYP3A enzymes. In particular, the activation of pregnane X receptor (PXR) by prototypical inducers of CYP3A has been found to correlate well with the species-specific modulation of CYP3A by various drugs and other xenobiotics. Concomitant with the discovery of PXR has been the identification of compounds that may act synergistically or antagonistically on multiple receptors (e.g., co-repressors and/or co-activators of the receptor) introducing novel mechanisms of drug-drug interactions. Differential expression of the individual isoforms of the major CYP450 enzymes over time in culture suggest that this model system is not reflective of in vivo profiles and, therefore, may be limited in its application for drug metabolism studies. Overall, primary cultures of human hepatocytes can serve as a sensitive and selective model for predicting the regulation of CYP450 modulation by drugs and other xenobiotics. Considerations and recommendations for standardizing testing conditions and choosing relevant endpoint(s) are presented.

Animals↗