SEARCH · PubMed Health

Results for “Model interpretability”

Explore indexed PubMed citations for clinical trials, systematic reviews and public health research. Read source abstracts and follow each citation to its original PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 775 records · Page 43Linked to original sources

PubMed2003

Bayesian latent variable models for median regression on multiple outcomes.

Often a response of interest cannot be measured directly and it is necessary to rely on multiple surrogates, which can be assumed to be conditionally independent given the latent response and observed covariates. Latent response models typically assume that residual densities are Gaussian. This article proposes a Bayesian median regression modeling approach, which avoids parametric assumptions about residual densities by relying on an approximation based on quantiles. To accommodate within-subject dependency, the quantile response categories of the surrogate outcomes are related to underlying normal variables, which depend on a latent normal response. This underlying Gaussian covariance structure simplifies interpretation and model fitting, without restricting the marginal densities of the surrogate outcomes. A Markov chain Monte Carlo algorithm is proposed for posterior computation, and the methods are applied to single-cell electrophoresis (comet assay) data from a genetic toxicology study.

Algorithms↗

PubMedDate not supplied

[Modeling real eukaryotic control gene subnetworks based on generalized threshold models].

Mathematical and computational means are developed that take into consideration the specifics of control processes at the molecular level and allow one to obtain both qualitative and quantitative patterns of gene network dynamics. Using the method of generalized threshold models, models are constructed for the Arabidopsis thaliana flower morphogenesis control subsystem and gene subnetwork controlling the Drosophila melanogaster early ontogeny. The dynamics of these systems are investigated: kinetic curves are computed for molecular components (RNA, proteins), possible modes of functioning and steady states of the nets are revealed and biologically interpreted. The models are shown to be adequate to the real processes. The effectiveness of the generalized threshold model method is evaluated in the analysis of the actual eukaryotic gene networks.

Animals↗

PubMed2004

Interpretation of normative thyroid volumes in children and adolescents: is there a need for a multivariate model?

The use of thyroid ultrasonography for determination of thyroid volume requires reliable reference criteria. The current World Health Organization/International Council for the Control of Iodine Deficiency Disorders (WHO/ICCIDD) reference has been questioned since iodine-sufficient children have been found throughout the world with distinctly smaller thyroid volumes. A difference in part explained by a systematic bias when generating the WHO/ICCIDD reference data. The objective with this study was to evaluate normative thyroid volumes in our region and, if possible, develop a multivariate model for their interpretation. Thyroid ultrasonography was performed and anthropometrical measurements were taken in 561 children and adolescents. The best predictor for thyroid volume in both girls and boys was body surface area (BSA), followed by age, weight, and height. References for normative thyroid volumes were calculated for each of the predictors. When these references were compared with other references throughout the world, the age-specific references were in good accord, but distinct differences were found between our BSA-specific references and other references based on a majority of children younger than ours. Using multivariate analyses, BSA and age were found to significantly influence thyroid volume, independently of each other. Multiple regression models by gender using BSA and age as predictors for thyroid volume were constructed. Using these models the difference between the BSA-specific references could be markedly reduced. To interpret thyroid volume accurately we propose the use of a multivariate model using age and BSA as predictors of thyroid volume.

Adolescent↗

PubMed1992

A coupled-oscillator model of ovarian-cycle synchrony among female rats.

The ovarian cycles of female rats become synchronized when they live together, as do the cycles of many other mammals. Ovarian cycles also become synchronized when rats live apart if they share a common air supply, indicating that ovarian-cycle synchrony is mediated by pheromones. We developed a coupled-oscillator model of ovarian-cycle synchrony to test several hypotheses about its pheromonal and neuroendocrine mechanisms and to guide our experimental research. The model spans three levels of organization: the group, the rat, and the neuroendocrine components of the ovarian system. The ovarian system (not the ovaries themselves) are modeled as an oscillating system. Coupling among ovarian systems is mediated by the exchange of two pheromones, one that delays the phase of the ovarian system and one that advances it. Computer simulation experiments showed that this coupled-oscillator model can explain the levels of ovarian-cycle synchrony observed in groups of female rats while, at the same time, matching an empirical distribution of ovarian-cycle lengths. By successfully matching computer simulation data with empirical data, we were able to infer theoretical predictions in a number of areas: (1) effect of initial conditions on the probability that a group will change to different synchrony level and phase relationships, i.e. the transition probability between all synchrony levels and phase relationships; (2) effects of individual differences in pheromone sensitivity on ovarian-cycle synchrony; (3) the timing of pheromone sensitivity during the ovarian cycle; and (4) the existence of partial luteinizing hormone surges, which may cause the "spontaneous" prolonged ovarian cycles associated with ovarian-cycle synchrony. The paper concludes by discussing the integrative role of this model for experimental research. In particular, we focus on the role of this model in interpreting theoretical aspects of ovarian-cycle synchrony as well as for guiding future experimental research into its mechanisms and functions.

Animals↗

PubMed2005

Stokes parameter studies of spontaneous emission from chiral nematic liquid crystals as a one-dimensional photonic stopband crystal: experiment and theory.

The helical structure of uniformly aligned chiral nematic liquid crystals results in a photonic stopband for only one sense of circular polarization. The spectroscopic Stokes polarimeter is used to analyze spontaneous emission in the stopband. Highly polarized photoluminescence is found and the polarization properties vary with the excitation wavelength. Spontaneous emission is enhanced at the stopband edge and this Purcell effect is greater on excitation at wavelengths where the absorption coefficient is low. This is interpreted as greater overlap between the excited molecules and the electrical modal field of the resonant modes at the stopband edge. Photoluminescence detected from the excitation face of the liquid crystal cell is less polarized because of photon tunneling. Fermi's golden rule in conjunction with Stokes vectors is used to model the polarization of emission taking multiple reflections at the interfaces of the cell into account. The discrepancy between the experiment and the theoretical model is interpreted as direct experimental evidence that virtual photons, which originate from zero point fluctuations of quantum space, are randomly polarized.

Journal Article↗

PubMed1994

Carbon isotope effects associated with aceticlastic methanogenesis.

The carbon isotope effects associated with synthesis of methane from acetate have been determined for Methanosarcina barkeri 227 and for methanogenic archaea in sediments of Wintergreen Lake, Michigan. At 37 degrees C, the 13C isotope effect for the reaction acetate (methyl carbon) --> methane, as measured in replicate experiments with M. barkeri, was - 21.3% +/- 0.3%. The isotope effect at the carboxyl portion of acetate was essentially equal, indicating participation of both positions in the rate-determining step, as expected for reactions catalyzed by carbon monoxide dehydrogenase. A similar isotope effect, - 19.2% +/- 0.3% was found for this reaction in the natural community (temperature = 20 degrees C). Given these observations, it has been possible to model the flow of carbon to methane within lake sediment communities and to account for carbon isotope compositions of evolving methane. Extension of the model allows interpretation of seasonal fluctuations in 13C contents of methane in other systems.

Acetates↗

PubMed2000

Genomic fold assignment and rational modeling of proteins of biological interest.

The first available genome of a multicellular organism, C. elegans, was used as a test case for protein fold assignment using PSI-BLAST, followed by rational structure modeling and interpretation of experimental mutagenesis data in the context of collaboration with biologists. Similar results are demonstrated for human disease proteins with known polymorphisms.

Animals↗

PubMed2026

An encyclopedia of human enhancer-gene regulatory interactions.

Identifying transcriptional enhancers and their target genes is essential for understanding gene regulation and the effect of human genetic variation on disease1-6. Here we create and evaluate a resource of more than 92 million enhancer-gene regulatory interactions across 1,458 biosamples covering 369 cell types and tissues, by integrating predictive models, chromatin states, three-dimensional contacts and large-scale genetic perturbations generated by the ENCODE Consortium7. We first create a systematic benchmarking pipeline to compare predictive models, assembling a dataset of 10,356 element-gene pairs measured in CRISPR perturbation experiments, more than 30,000 fine-mapped expression quantitative trait loci and 569 fine-mapped genome-wide association study (GWAS) variants linked to a probable causal gene. Using this framework, we develop ENCODE-rE2G, a predictive model achieving state-of-the-art performance across several prediction tasks, demonstrating that iterative perturbations and supervised machine learning can build increasingly accurate predictive models of enhancer regulation. Using ENCODE-rE2G, we build an encyclopedia of enhancer-gene regulatory interactions in the human genome, revealing global properties of enhancer networks, identifying differences in regulatory complexity across genes and improving analyses linking noncoding variants to target genes and cell types for common complex diseases. By interpreting the model, we find that beyond enhancer activity and three-dimensional enhancer-promoter contacts, additional features that guide enhancer-promoter communication include promoter class and enhancer-enhancer synergy. These genome-wide maps of enhancer-gene regulatory interactions, benchmarking software, predictive models and insights about enhancer function provide a valuable resource for future studies of gene regulation and human genetics.

Humans↗

PubMed1993

Exploring the relationship between rationality and bounded rationality in medical knowledge-based systems.

If our goal in Artificial Intelligence in Medicine (AIM) is to engineer systems health-care providers will both use and, in the process, improve their performance, we must concentrate on the development of causal theories of knowledge and problem solving. One broad direction in pursuing this goal is understanding the relationships between existing models of rationality and bounded rationality for similar tasks. Models of rationality refer to those approaches in which the optimal properties of the models are deductively provable, i.e. in which the processing is rational. Representative models of rationality used in AIM are deductive logical models, statistical models such as Bayesian inference models, and decision-analytic models. Models of bounded rationality are those which do not guarantee such optimal properties nor yield to deductive correctness proofs. These models have their roots in cognitive psychology. In this article we show how explicating the relationship between models of rationality and bounded rationality might be done in the case of abductive tasks in medicine. This is done by positioning these modeling approaches within the same framework (an abstract computational model) and interpreting in this context both computational complexity results concerning the nature of the task and empirical results studies of human problem-solving behavior.

Artificial Intelligence↗

PubMed1997

The lasso method for variable selection in the Cox model.

I propose a new method for variable selection and shrinkage in Cox's proportional hazards model. My proposal minimizes the log partial likelihood subject to the sum of the absolute values of the parameters being bounded by a constant. Because of the nature of this constraint, it shrinks coefficients and produces some coefficients that are exactly zero. As a result it reduces the estimation variance while providing an interpretable final model. The method is a variation of the 'lasso' proposal of Tibshirani, designed for the linear regression context. Simulations indicate that the lasso can be more accurate than stepwise selection in this setting.

Humans↗

PubMed2004

A model of the dynamic relationship between blood flow and volume changes during brain activation.

The temporal relationship between changes in cerebral blood flow (CBF) and cerebral blood volume (CBV) is important in the biophysical modeling and interpretation of the hemodynamic response to activation, particularly in the context of magnetic resonance imaging and the blood oxygen level-dependent signal. measured the steady state relationship between changes in CBV and CBF after hypercapnic challenge. The relationship CBV is proportional to CBFphi has been used extensively in the literature. Two similar models, the Balloon and the Windkessel , have been proposed to describe the temporal dynamics of changes in CBV with respect to changes in CBF. In this study, a dynamic model extending the Windkessel model by incorporating delayed compliance is presented. The extended model is better able to capture the dynamics of CBV changes after changes in CBF, particularly in the return-to-baseline stages of the response.

Animals↗

PubMed2004

Modelling land use change with generalized linear models--a multi-model analysis of change between 1860 and 2000 in Gallatin Valley, Montana.

This paper develops an approach to modelling land use change that links model selection and multi-model inference with empirical models and GIS. Land use change is frequently studied, and understanding gained, through a process of modelling that is an empirical analysis of documented changes in land cover or land use patterns. The approach here is based on analysis and comparison of multiple models of land use patterns using model selection and multi-model inference. The approach is illustrated with a case study of rural housing as it has developed for part of Gallatin County, Montana, USA. A GIS contains the location of rural housing on a yearly basis from 1860 to 2000. The database also documents a variety of environmental and socio-economic conditions. A general model of settlement development describes the evolution of drivers of land use change and their impacts in the region. This model is used to develop a series of different models reflecting drivers of change at different periods in the history of the study area. These period specific models represent a series of multiple working hypotheses describing (a) the effects of spatial variables as a representation of social, economic and environmental drivers of land use change, and (b) temporal changes in the effects of the spatial variables as the drivers of change evolve over time. Logistic regression is used to calibrate and interpret these models and the models are then compared and evaluated with model selection techniques. Results show that different models are 'best' for the different periods. The different models for different periods demonstrate that models are not invariant over time which presents challenges for validation and testing of empirical models. The research demonstrates (i) model selection as a mechanism for rating among many plausible models that describe land cover or land use patterns, (ii) inference from a set of models rather than from a single model, (iii) that models can be developed based on hypothesised relationships based on consideration of underlying and proximate causes of change, and (iv) that models are not invariant over time.

Agriculture↗

PubMed2003

Mapping property distributions of molecular surfaces: algorithm and evaluation of a novel 3D quantitative structure-activity relationship technique.

A novel molecular descriptor called MaP (mapping property distributions of molecular surfaces) is presented. It combines facile computation, translational and rotational invariance, and straightforward interpretability of the computed models. A three-step procedure is used to compute the MaP descriptor. First, an approximation to the molecular surface with equally distributed surface points is computed. Next, molecular properties are projected onto this surface. Finally, the distribution of surface properties is encoded into a translationally and rotationally invariant molecular descriptor that is based on radial distribution functions (distance-dependent count statistics). The calculated descriptor is correlated with biological data through chemometric regression techniques in combination with a variable selection. The latter is used to identify variables that are highly relevant for the model and hence for its interpretation. Three applications of the new descriptor are presented, each representing a different area of 3D-QSAR. For reasons of comparability, the new descriptor was tested on the steroid "benchmark" data set. Furthermore, a highly diverse data set with potentially eye-irritating compounds was studied, and third, a set of flexible structures with a modulating effect on the muscarinic M(2) receptor were studied. Not only were all models highly predictive but interpretation of the back-projected variables into the original molecular space led to biologically and chemically relevant conclusions.

Acetates↗

PubMed1983

A model for the temperature distribution in skin noxiously stimulated by a brief pulse of CO2 laser radiation.

The application of localized noxious heat stimuli to the skin, generated by brief infrared radiation pulses emitted by a CO2 laser, is a relatively new experimental technique for the thermal induction of pain in humans and in experimental animals. This study proposes a model for the spatial (3-dimensional) and temporal distribution of the skin temperature during and following a radiation pulse. The heat equation is written and solved, using thermal and optical constants of human skin reported in the literature. The solution is approximated, with a very small error, by a closed form expression, having a simple physical interpretation. This model is applied to analyze a typical set-up currently in use in our laboratory. The results show a significant difference between the temperature of the surface of the skin and that of the border between the epidermis and the dermis, which is the location of the most superficial receptive nerve ends. It is shown that, for the set-up examined, these nerve ends reach a temperature of 45 degrees C, known to be the human pain threshold, 30-40 ms after pulse onset. Moreover, it is also shown that they may remain above threshold temperature for up to a few hundreds of milliseconds (considerably outlasting pulse cessation). In addition, it is shown that the area in which nerve ends reach this threshold is a circle with a very small radius (1-2.5 mm). The implications of the results on the double sensation experienced by humans, and on the extremely powerful EEG correlates, are discussed.

Humans↗

PubMed1997

A neural model of multimodal adaptive saccadic eye movement control by superior colliculus.

How does the saccadic movement system select a target when visual, auditory, and planned movement commands differ? How do retinal, head-centered, and motor error coordinates interact during the selection process? Recent data on superior colliculus (SC) reveal a spreading wave of activation across buildup cells the peak activity of which covaries with the current gaze error. In contrast, the locus of peak activity remains constant at burst cells, whereas their activity level decays with residual gaze error. A neural model answers these questions and simulates burst and buildup responses in visual, overlap, memory, and gap tasks. The model also simulates data on multimodal enhancement and suppression of activity in the deeper SC layers and suggests a functional role for NMDA receptors in this region. In particular, the model suggests how auditory and planned saccadic target positions become aligned and compete with visually reactive target positions to select a movement command. For this to occur, a transformation between auditory and planned head-centered representations and a retinotopic target representation is learned. Burst cells in the model generate teaching signals to the spreading wave layer. Spreading waves are produced by corollary discharges that render planned and visually reactive targets dimensionally consistent and enable them to compete for attention to generate a movement command in motor error coordinates. The attentional selection process also helps to stabilize the map-learning process. The model functionally interprets cells in the superior colliculus, frontal eye field, parietal cortex, mesencephalic reticular formation, paramedian pontine reticular formation, and substantia nigra pars reticulata.

Animals↗

PubMed1998

Ocular fixation index and mathematical models.

Ocular fixation test and ocular fixation index (OFI) never have been interpreted in terms of mathematical models, despite their widespread diffusion. However, ocular fixation is a typical case of visual-vestibular interaction, and mathematical models have proven very helpful in interpreting some mechanisms of this interaction, e.g. those of the optokinetic-vestibular interaction. In the present paper, a first attempt is proposed toward a model interpretation of OFI. By using very simple mathematical models, the hypothesis is tested that visual suppression of vestibular nystagmus results from direct action of smooth pursuit system (SPS). The aim is to draw consequences and recognize possible limits of this hypothesis. Dependence of OFI on SPS performance is examined. Although the available experimental data are insufficient for comprehensive validation of the model, the results agree with the current interpretations. In particular, quantitative support is given to the sensitivity of OFI to central vestibular diseases. Although the interpretation of visual suppression and OFI in terms of mathematical models is still at a very preliminary stage, models may provide a theoretical reference framework for the interpretation of new experimental results and/or suggest new test protocols.

Fixation, Ocular↗

PubMed1997

Anisotropic molecular rotational diffusion in 15N spin relaxation studies of protein mobility.

The backbone dynamics of the uniformly 15N-labeled N-terminal 63-residue DNA-binding domain of the 434 repressor has been characterized by measurements of the individual 15N longitudinal relaxation times, T1, transverse relaxation times, T2, and heteronuclear 15N[1H]-NOEs at 1H resonance frequencies of 400 and 750 MHz. The dependence of an apparent spherical top correlation time, tauR, on the orientation of the N-H bond vector with respect to the principal axes of the global diffusion tensor of the protein was used to establish the fact that the degree of anisotropy of the global molecular tumbling amounts to 1.2, which is in good agreement with the values obtained from model calculations of the hydrodynamic properties. A model-free analysis showed that even this small anisotropy leads to the implication of artifactual slow internal motions for at least two residues when the assumption of isotropic global motion is used. Additional residues may actually undergo internal motions on the same time scale as the global rotational diffusion, in which case the model-free approach would, however, be inappropriate for quantifying the correlation times and order parameters. Overall, the experiments with 434(1-63) demonstrate that the assumption of isotropic rotational reorientation may result in artifacts of model-free interpretations of spin relaxation data even for proteins with small deviations from spherical shape.

Binding Sites↗

PubMed1993

Circulatory models of intact-body kinetics and their relationship with compartmental and non-compartmental analysis.

Circulatory models for interpreting the kinetics of substances in vivo explain the kinetics of the intact body with the concepts of organ kinetics. In this paper, it is first shown that classical organ kinetic analysis is incomplete for metabolized substances, and an appropriate extension is developed. It is then discussed how the concepts of the extended organ kinetic analysis apply to circulatory models. At an organ level, it is shown that two impulse responses are necessary to characterize the organ, one relating influx and outflux, and one relating influx and uptake. The consideration of the inlet-uptake path is of fundamental importance for a correct calculation of the organ volume. It is demonstrated that the total volume is the sum of two components, the first related to the inlet-outlet path, the second to the inlet-uptake path. The first term is computable without assumptions, while the second is model-dependent. Analogous results hold at a total-body level. The relationships between circulatory models and compartmental and non-compartmental analysis are also precisely established. The significance of the non-compartmental estimate of the distribution volume is clarified. The usual strategy of compartmental modeling by which losses are placed in peripheral compartments according to a correspondence between tissues and compartments is shown to be misleading. An example concerning glucose kinetics is given. In conclusion, this paper shows that many of the dominating paradigms of organ and total-body kinetic analysis must be revised.

Humans↗