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 595 records · Page 33Linked to original sources

PubMedDate not supplied

The IUPS Physiome Project: a framework for computational physiology.

The IUPS Physiome Project is an internationally collaborative open-source project to provide a public domain framework for computational physiology, including the development of modelling standards, computational tools and web-accessible databases of models of structure and function at all spatial scales. A number of papers in this volume deal with the development of specific mathematical models of physiological processes. This paper stands back from the detail of individual models and reviews the current state of the IUPS Physiome Project including organ and organ system continuum models, the interpretation of constitutive law parameters in terms of micro-structural models, and markup languages for standardizing cellular processes. Some current practical applications of the physiome models are given and some of the challenges for the next 5 years of the Physiome Project at the level of organs, cells and proteins are proposed.

Cell Physiological Phenomena↗

PubMed2005

The influence of the rigidity of geosorbent organic matter on non-ideal sorption behaviors of chlorinated benzenes.

This study focused on evaluating the influence of the rigidity of natural organic matter (NOM) associated with four natural geosorbents in controlling the non-ideal sorption behaviors of five chlorinated benzenes. Single solute sorption isotherms for each sorbate/sorbent combination were modeled and interpreted by the Freundlich sorption isotherm and the adsorption-partitioning model based on Polanyi-Manes theory (PM model). "Rigid" organic matter was operationally quantified as the fraction of carbon resistant to wet chemical oxidation (hard carbon) or thermal oxidation (soot carbon); atomic H/O ratios indicated a close correlation between the degree of reduction of the NOM and its rigidity. Sorbents with larger rigid carbon fractions had more non-linear sorption isotherms and higher organic carbon (OC) normalized sorption affinities. The size of the PM hole filling domain for a given sorbent was closely correlated with the extent to which the sorbent's affinity for chlorobenzenes exceeded predictions from a linear free energy relationship. Loss of some portions of the rigid character of the NOM domain due to the penetration of sorbate molecules (plasticization) was discussed as a possible contributor to the non-ideal sorption behaviors observed in this study. The existence of a permanently rigid NOM domain, not subject to plasticization under environmental conditions, was postulated as an additional factor determining the observed sorption behavior.

Adsorption↗

PubMed2000

Electroencephalographic measurement of motor cortex control of muscle activity in humans.

OBJECTIVE: To detect and measure correlation between cortical and muscle activities, coherence analysis was used. METHODS: The electroencephalogram (EEG) and electromyogram (EMG) were recorded in 9 normal volunteers during tonic contraction of upper and lower limb muscles on the right side. Coherence between EEG and EMG was computed to analyze their linear association. RESULTS: EEG over the contralateral sensorimotor area was coherent with EMG, with peak coherence at 11-36 Hz (mean, 22 Hz). For the abductor pollicis brevis (APB) muscle, peak coherence, as determined by functional brain mapping with focal transcranial magnetic stimulation (TMS), was over or slightly posterior to the hand area on the primary motor cortex determined by focal transcranial magnetic stimulation (TMS). Peak coherence over the scalp was somatotopically organized. The temporal relation between EEG and EMG was analyzed with a new model for interpreting the phase shift ('constant phase shift plus constant time lag' model). For the APB muscle, the phase relation between cortical and muscular oscillations differed in the frequency ranges of 3-13 Hz and 14-50 Hz, respectively, suggesting that different coupling mechanisms operate in different bands. Only the phase shift between cortical and motoneuronal firing at 14-50 Hz was reliably estimated by a linear model. At 14-50 Hz, motoneuronal firing was led by surface-negative cortical activity with a constant time lag that depended on the cortical-muscular distance. For the APB muscle, the time lag was slightly shorter than the cortical-muscular conduction time determined by TMS. Vibratory stimulation (100 Hz) of a muscle tendon during tonic contraction had no significant effect on cortical-muscular coherence, indicating that cortical oscillation reflected motor rather than sensory activity. CONCLUSIONS: The present findings suggest temporal coding of the oscillatory motor control system (3-13 Hz vs. 14-50 Hz), and confirm the functional importance of cortical beta and gamma rhythms in the motor efferent command. Cortical-muscular synchronization is most likely mediated by the direct corticospinal pathway within the frequency range of 14-50 Hz.

Adult↗

PubMedDate not supplied

Ensemble of linear models for predicting drug properties.

We propose a new classification method for the prediction of drug properties, called random feature subset boosting for linear discriminant analysis (LDA). The main novelty of this method is the ability to overcome the problems with constructing ensembles of linear discriminant models based on generalized eigenvectors of covariance matrices. Such linear models are popular in building classification-based structure-activity relationships. The introduction of ensembles of LDA models allows for an analysis of more complex problems than by using single LDA, for example, those involving multiple mechanisms of action. Using four data sets, we show experimentally that the method is competitive with other recently studied chemoinformatic methods, including support vector machines and models based on decision trees. We present an easy scheme for interpreting the model despite its apparent sophistication. We also outline theoretical evidence as to why, contrary to the conventional AdaBoost ensemble algorithm, this method is able to increase the accuracy of LDA models.

ATP Binding Cassette Transporter, Subfamily B, Mem↗

PubMed2002

An integrative model of coupled water and solute exchange in the heart.

Physiologists have devised many models for interpreting water and solute exchange data in whole organs, but the models have typically neglected key aspects of the underlying physiology to present the simplest possible model for a given experimental situation. We have developed a physiologically realistic model of microcirculatory water and solute exchange and applied it to diverse observations of water and solute exchange in the heart. Model simulations are consistent with the results of osmotic weight transient, tracer indicator dilution, and steady-state lymph sampling experiments. The key model features that permit this unification are the use of an axially distributed blood-tissue exchange region, inclusion of a lymphatic drain in the interstitium, and the independent computation of transcapillary solute and solvent fluxes through three different pathways.

Animals↗

PubMed2003

Murine models of asthma.

In vivo animal models can offer valuable information on several aspects of asthma pathogenesis and treatment. The mouse is increasingly used in these models, mainly because this species allows for the application in vivo of a broad range of immunological tools, including gene deletion technology. Mice, therefore, seem particularly useful to further elucidate factors influencing the response to inhaled allergens. Examples include: the role of immunoregulatory mechanisms that protect against T-helper cell type 2 cell development; the trafficking of T-cells; and the contribution of the innate immunity. However, as for other animal species, murine models also have limitations. Mice do not spontaneously develop asthma and no model mimics the entire asthma phenotype. Instead, mice should be used to model specific traits of the human disease. The present task force report draws attention to specific aspects of lung structure and function that need to be borne in mind when developing such models and interpreting the results. In particular, efforts should be made to develop models that mimic the lung function changes characteristic of asthma as closely as possible. A large section of this report is therefore devoted to an overview of airway function and its measurement in mice.

Animals↗

PubMed1998

Rodent models of cardiopulmonary disease: their potential applicability in studies of air pollutant susceptibility.

The mechanisms by which increased mortality and morbidity occur in individuals with preexistent cardiopulmonary disease following acute episodes of air pollution are unknown. Studies involving air pollution effects on animal models of human cardiopulmonary diseases are both infrequent and difficult to interpret. Such models are, however, extensively used in studies of disease pathogenesis. Primarily they comprise those developed by genetic, pharmacologic, or surgical manipulations of the cardiopulmonary system. This review attempts a comprehensive description of rodent cardiopulmonary disease models in the context of their potential application to susceptibility studies of air pollutants regardless of whether the models have been previously used for such studies. The pulmonary disease models include bronchitis, emphysema, asthma/allergy, chronic obstructive pulmonary disease, interstitial fibrosis, and infection. The models of systemic hypertension and congestive heart failure include: those derived by genetics (spontaneously hypertensive, Dahl S. renin transgenic, and other rodent models); congestive heart failure models derived by surgical manipulations; viral myocarditis; and cardiomyopathy induced by adriamycin. The characteristic pathogenic features critical to understanding the susceptibility to inhaled toxicants are described. It is anticipated that this review will provide a ready reference for the selection of appropriate rodent models of cardiopulmonary diseases and identify not only their pathobiologic similarities and/or differences to humans but also their potential usefulness in susceptibility studies.

Air Pollutants↗

PubMed2026

Assessing Metal Ion Assignment Accuracy in Protein Data Bank Models via Elemental Spectroscopy.

Accurate representation of metal ions in macromolecular structures is critical for chemical interpretation, computational modeling, and machine-learning methods that rely on Protein Data Bank (PDB) entries. However, the elemental identity of metals modeled in crystallographic structures is often inferred indirectly and rarely validated experimentally. Here, we combine Particle Induced X-ray Emission (PIXE) and X-ray Fluorescence Spectroscopy (XRFS) to determine the elemental composition of protein samples used to generate 70 deposited metalloprotein crystal structures. By analyzing the original protein material employed for crystallization, but before the addition of crystallization buffer solutions, we assess whether the modeled metal ions in deposited structures are consistent with experimentally detectable elemental content. We find that in a majority of cases, the metals modeled in the corresponding PDB entries are inconsistent with the metals present in the protein samples before crystallization, or that additional metals are present but not represented in the structural models. Spectroscopic results were integrated with automated crystallographic validation metrics, including real-space Z-difference (RSZD) analysis and systematic rerefinement, to evaluate atomic-number mismatch at metal sites. PIXE and XRFS show strong agreement for dominant elemental signals and provide complementary, scalable approaches for identifying suspect metal assignments. This work does not address physiological or functional metalation but instead highlights a widespread data integrity issue in deposited macromolecular structures, PDB-wide. These results establish an experimentally corroborated link between elemental identity and crystallographic validation metrics, enabling the large-scale detection of chemically inconsistent annotations in structural databases used for computational modeling and machine learning.

Databases, Protein↗

PubMed2004

Using models of the myocyte for functional interpretation of cardiac proteomic data.

There has been significant progress towards the development of highly integrative computational models of the cardiac myocyte over the past decade. Models now incorporate descriptions of voltage-gated ionic currents and membrane transporters, mechanisms of calcium-induced calcium release and intracellular calcium cycling, mitochondrial ATP production and its coupling to energy-requiring membrane transport processes and mechanisms of force generation. There is an extensive literature documenting both the reconstructive and predictive abilities of these models and there is no question that an interplay between quantitative modelling and experimental investigation has become a central component of modern cardiovascular research. As data regarding the cardiovascular proteome in both health and disease emerge, integrative models of the myocyte are becoming useful tools for interpreting the functional significance of changes in protein expression and post-translational modifications (PTMs). Data of particular importance include information on: (a) changes of expressed protein level, (b) changes of protein PTMs, (c) protein localization, and (d) protein-protein interactions, as it is often possible to incorporate and interpret the functional significance of such findings using computational models. We provide two examples of how models may be used in this fashion. In the first example, we show how information on altered expression of the sarcoplasmic reticulum Ca2+-ATPase, when interpreted through the use of a computational model, has provided key insights into fundamental mechanisms regulating cardiac action potential duration. In the second example, we show how information on the effects of phosphorylation of L-type Ca2+ channels, when interpreted through the use of a model, provides insights on how this post-translational modification alters the properties of excitation-contraction coupling and risk for arrhythmia.

Animals↗

PubMed1991

Analysis of failure time data with ordinal categories of response.

When failure times are observed, additional information concerning the type of failure is often recorded. A method which simultaneously models the failure times and additional information in the form of ordinal categories is discussed. An application to clinical trial data, in which the failure times are times of onset of headache, and the headaches are classified into the ordinal categories mild, moderate and severe, illustrates how this method may be used and how the final model can be interpreted. The continuation ratio model, which is used in this method, is described in detail.

Clinical Trials as Topic↗

PubMed2004

Retinal neurodegeneration in the DBA/2J mouse-a model for ocular hypertension.

Mice of the DBA/2J strain spontaneously develop complex ocular abnormalities, including glaucomatous loss of retinal ganglion cells (RGC). In the present study ultrastructural features of retinal neurodegeneration in DBA/2J mice of different age (3, 6, 8 and 11 months) are described. By 3 months, RGC apoptosis characterized by electron-dense karioplasm and cytoplasm of ganglion cells was observed. The occurrence of apoptotic ganglion cells peaked at the age of 6 months. Past this age, necrosis characterized by swelling and electron-rare cytoplasm appeared to be the prevailing form of cell death. Müller glia activation increased with age, but there were no signs of leukocyte infiltration. At 8 and 11 months, signs of neoangiogenesis were found both at the ultrastructural level and in clinical examinations. In these older animals myelin-like bodies, most probably representing the intracellular aggregates of phospholipids in irreversibly injured cells, were also seen. Photoreceptor cells were not affected at any age. Our observations suggest that retinal degeneration in the DBA/2J mice does not involve recruitment of blood-borne inflammatory/phagocytosing cells, and that apoptosis is gradually replaced by necrosis as the predominant pathway of RGC death. Retinal degeneration in 3- to 11-month-old DBA/2J mice partially resembles human pigment dispersion syndrome and pigmentary glaucoma with characteristic anterior segment changes and elevation of intraocular pressure. However, neovasculogenesis and myelin-like bodies are observed during aging. Therefore, the DBA/2J model requires judicious interpretation as a glaucoma model.

Age Factors↗

PubMed1998

A little to a lot or a lot to a little: is NTCP always minimized in multiport therapy?

PURPOSE: We address the question of whether or not, for the same average (or integral) dose, a smaller uniform dose to an entire normal tissue structure always results in a lower normal tissue complication probability (NTCP) than does a proportionally larger dose to a partial volume of the same structure. METHODS AND MATERIALS: A recent compilation of NTCP data and two theoretical formulations of the dependence of NTCP on dose and partial volume irradiated-the Lyman probit equation and the binomial model-are used to examine this question. Both models fit equally well available NTCP data. RESULTS: Empirical data indicate that for lung, kidney, and possibly liver (but not for esophagus, brain, or heart), given a fixed tumor dose and fixed integral dose, NTCP can be minimized by irradiating a partial volume fraction rather than the entire normal organ. The binomial model supports this interpretation, whereas the probit model predicts that for all organs uniform irradiation of the whole organ always results in the lowest possible NTCP. CONCLUSIONS: In contrast to what is commonly believed, this study suggests that for at least two normal tissues, namely lung and kidney, there may be situations where "a lot to a little" (i.e., fewer treatment ports) will result in higher tumor control probability and better treatment plan than "a little to a lot" (i.e., multifield treatment). This finding, which is independent of the binomial or probit models used here, depends only on the accuracy of the empirical NTCP data. It is also interesting to note that: a) lung and kidney are commonly classified as parallel tissues, while the others have more of a serial architecture; and b) the choice of the NTCP model can have a profound impact on treatment planning decisions.

Models, Theoretical↗

PubMed2003

A simulation study of the error in dipole source localization for EEG spikes with a realistic head model.

OBJECTIVE: We tried to determine the error range of dipole modeling for EEG spikes originating from various clinically important sources by a simulation study employing a realistic head model. The computed error range was also compared with the degree of disturbance of dipole modeling caused by adding background activity to the spike. METHODS: The scalp fields generated by temporal, frontal and rolandic epileptic sources with spatial extent were simulated, and the corresponding 3-dimensional maps of residual variance (RV) were built by computing the RV for a single dipole at each point on a fine imaginary grid in the brain. Single dipole modeling was also performed for the simulated scalp fields after adding real background activity. RESULTS: The brain volume corresponding to a small RV was compact for the frontal sources and the lateral and baso-mesial temporal sources, and large for the anterior and baso-lateral temporal sources. The distribution of dipoles estimated for spikes contaminated with background corresponded to that of the volume of small RV and to spike-amplitude. Estimates were improved by employing inferior temporal electrodes. CONCLUSIONS: When evaluating dipole models of epileptic spikes, error ranges can be estimated and they vary considerably from region to region. SIGNIFICANCE: This study illustrates the variability of the error in dipole modeling of epileptic spikes. This variability is important when considering the clinical interpretation of modeling results.

Brain Mapping↗

PubMed2002

Modelling viral and immune system dynamics.

During the past 6 years, there have been substantial advances in our understanding of human immunodeficiency virus 1 and other viruses, such as hepatitis B virus and hepatitis C virus, that cause chronic infection. The use of mathematical modelling to interpret experimental results has made a significant contribution to this field. Mathematical modelling is also improving our understanding of T-cell dynamics and the quantitative events that underlie the immune response to pathogens.

Animals↗

PubMed2004

Reaction of Cu+ with dimethoxyethane: competition between association and multiple dissociation channels.

The reaction of Cu+ with dimethoxyethane (DXE) is studied using kinetic-energy dependent guided ion beam mass spectrometry. The bimolecular reaction forms an associative Cu(+)(DXE) complex that is long-lived and dissociates into several competitive channels: C4H9O2(+)+CuH, Cu(+)(C3H6O)+CH3OH, back to reactants, and other minor channels. The kinetic-energy dependences of the cross sections for the three largest product channels are interpreted with several different models (including rigorous phase space theory) to yield 0 K bond energies after accounting for the effects of multiple ion-molecule collisions, internal energy of the reactant ions, Doppler broadening, and dissociation lifetimes. These values are compared with bond energies obtained from collision-induced dissociation (CID) studies of the Cu(+)(DXE) complex and found to be self-consistent. Although all models provide reasonable thermochemistry, phase space theory reproduces the details of the cross sections most accurately. We also examine the dynamics of this reaction using time-of-flight methods and a retarding potential analysis. This provides additional insight into the unimolecular decay of the long-lived Cu(+)(DXE) association complex. Comparison of results from this study with those from the complementary CID study, thus forming the same energized Cu(+)(DXE) complex in two distinct ways, allows an assessment of the models used to interpret CID thresholds.

Journal Article↗

PubMed1999

The use of fractional polynomials to model continuous risk variables in epidemiology.

BACKGROUND: The traditional method of analysing continuous or ordinal risk factors by categorization or linear models may be improved. METHODS: We propose an approach based on transformation and fractional polynomials which yields simple regression models with interpretable curves. We suggest a way of presenting the results from such models which involves tabulating the risks estimated from the model at convenient values of the risk factor. We discuss how to incorporate several continuous risk and confounding variables within a single model. The approach is exemplified with data from the Whitehall I study of British Civil Servants. We discuss the approach in relation to categorization and non-parametric regression models. RESULTS: We show that non-linear risk models fit the data better than linear models. We discuss the difficulties introduced by categorization and the advantages of the new approach. CONCLUSIONS: Our approach based on fractional polynomials should be considered as an important alternative to the traditional approaches for the analysis of continuous variables in epidemiological studies.

Coronary Disease↗

PubMed2002

Venusian "hot spots": physical phenomenon and its quantification.

An overall picture of the Venusian hot spots phenomenon is considered in the framework of the simplest conceptual models that admit the solutions in the form of steadily rotating "hot" vortices. Model assumptions take into account only those features of the middle atmosphere in the polar region of Venus that are supported by observational data and are essential for understanding the physical mechanism initiating similar vortices. The problem is analyzed in the framework of both the pointlike and petal-like models of cyclostrophic vortices. Interpretation of these models as an upper and lower bound of a complete theory allows one to find the region of existence of the regimes responsible for the Venusian hot spots and also to establish and assess numerically conditions under which such vortices can be formed. The emphasis is on a comparison of the theoretically established results with the observational data.

Journal Article↗

PubMed2002

A mixture model approach to indexing rater agreement.

Raters are an important potential source of measurement error when assigning targets to categories. Therefore, psychologists have devoted considerable attention to quantifying the extent to which ratings agree with each other. Two main approaches to analysing rater agreement data can be distinguished. While the first approach focuses on the development of summary statistics that index rater agreement, the second models the association pattern among the observers' ratings. With the modelling approach three groups of models can be distinguished: latent class models, simple quasisymmetric agreement models, and mixture models. This paper discusses a class of mixture models that is defined by its characteristic of having a quasi-symmetric log-linear representation. This class of models has two interesting properties. First, the simple quasi-symmetric agreement models can be shown to be members of this class. Therefore, the results of a rater agreement analysis based on a simple quasi-symmetric agreement model may be interpreted in the mixture model framework. Second, since the mixture models readily provide a familiar measure of rater reliability, it is possible to obtain a model-based estimate of rater reliability from the simple quasi-symmetric agreement models. The suggested class of mixture models will be illustrated using data from a persuasive communication study in which three raters classified respondents on the basis of their elicited cognitive responses.

Humans↗