We have developed a method to longitudinally classify subjects into two or more prognostic groups using longitudinally observed values of markers related to the prognosis. We assume the availability of a training data set where the subjects' allocation into the prognostic group is known. The proposed method proceeds in two steps as described earlier in the literature. First, multivariate linear mixed models are fitted in each prognostic group from the training data set to model the dependence of markers on time and on possibly other covariates. Second, fitted mixed models are used to develop a discrimination rule for future subjects. Our method improves upon existing approaches by relaxing the normality assumption of random effects in the underlying mixed models. Namely, we assume a heteroscedastic multivariate normal mixture for random effects. Inference is performed in the Bayesian framework using the Markov chain Monte Carlo methodology. Software has been written for the proposed method and it is freely available. The methodology is applied to data from the Dutch Primary Biliary Cirrhosis Study.
- MeSH
- Liver Cirrhosis, Biliary drug therapy MeSH
- Biomarkers analysis MeSH
- Cholagogues and Choleretics therapeutic use MeSH
- Discriminant Analysis MeSH
- Data Interpretation, Statistical MeSH
- Ursodeoxycholic Acid therapeutic use MeSH
- Humans MeSH
- Linear Models MeSH
- Longitudinal Studies MeSH
- Computer Simulation MeSH
- Disease Progression MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
- Research Support, Non-U.S. Gov't MeSH
Wiley series in probability and statistics
1st ed. xix, 506 s.
Functional connectivity (FC) analysis is a prominent approach to analyzing fMRI data, especially acquired under the resting state condition. The commonly used linear correlation FC measure bears an implicit assumption of Gaussianity of the dependence structure. If only the marginals, but not all the bivariate distributions are Gaussian, linear correlation consistently underestimates the strength of the dependence. To assess the suitability of linear correlation and the general potential of nonlinear FC measures, we present a framework for testing and estimating the deviation from Gaussianity by means of comparing mutual information in the data and its Gaussianized counterpart. We apply this method to 24 sessions of human resting state fMRI. For each session, matrix of connectivities between 90 anatomical parcel time series is computed using mutual information and compared to results from its multivariate Gaussian surrogate that conserves the correlations but cancels any nonlinearity. While the group-level tests confirmed non-Gaussianity in the FC, the quantitative assessment revealed that the portion of mutual information neglected by linear correlation is relatively minor-on average only about 5% of the mutual information already captured by the linear correlation. The marginality of the non-Gaussianity was confirmed in comparisons using clustering of the parcels-the disagreement between clustering obtained from mutual information and linear correlation was attributable to random error. We conclude that for this type of data, practical relevance of nonlinear methods trying to improve over linear correlation might be limited by the fact that the data are indeed almost Gaussian.
- MeSH
- Algorithms MeSH
- Adult MeSH
- Fourier Analysis MeSH
- Oxygen blood MeSH
- Humans MeSH
- Linear Models MeSH
- Magnetic Resonance Imaging MeSH
- Young Adult MeSH
- Neural Pathways physiology MeSH
- Normal Distribution MeSH
- Rest physiology MeSH
- Cluster Analysis MeSH
- Software MeSH
- Check Tag
- Adult MeSH
- Humans MeSH
- Young Adult MeSH
- Male MeSH
- Female MeSH
- Publication type
- Journal Article MeSH
- Research Support, Non-U.S. Gov't MeSH
Recently developed methods of longitudinal discriminant analysis allow for classification of subjects into prespecified prognostic groups using longitudinal history of both continuous and discrete biomarkers. The classification uses Bayesian estimates of the group membership probabilities for each prognostic group. These estimates are derived from a multivariate generalised linear mixed model of the biomarker's longitudinal evolution in each of the groups and can be updated each time new data is available for a patient, providing a dynamic (over time) allocation scheme. However, the precision of the estimated group probabilities differs for each patient and also over time. This precision can be assessed by looking at credible intervals for the group membership probabilities. In this paper, we propose a new allocation rule that incorporates credible intervals for use in context of a dynamic longitudinal discriminant analysis and show that this can decrease the number of false positives in a prognostic test, improving the positive predictive value. We also establish that by leaving some patients unclassified for a certain period, the classification accuracy of those patients who are classified can be improved, giving increased confidence to clinicians in their decision making. Finally, we show that determining a stopping rule dynamically can be more accurate than specifying a set time point at which to decide on a patient's status. We illustrate our methodology using data from patients with epilepsy and show how patients who fail to achieve adequate seizure control are more accurately identified using credible intervals compared to existing methods.
- MeSH
- Bayes Theorem * MeSH
- Discriminant Analysis MeSH
- Epilepsy diagnosis therapy MeSH
- Remission Induction MeSH
- Classification methods MeSH
- Humans MeSH
- Linear Models MeSH
- Longitudinal Studies MeSH
- Multivariate Analysis MeSH
- Computer Simulation MeSH
- Probability * MeSH
- Prognosis MeSH
- Decision Making MeSH
- Sensitivity and Specificity MeSH
- Check Tag
- Humans MeSH
- Publication type
- Journal Article MeSH
In comparison with analytical tools, bioassays provide higher sensitivity and more complex evaluation of environmental samples and are indispensable tools for monitoring increasing in anthropogenic pollution. Nevertheless, the disadvantage in cellular assays stems from the material variability used within the assays, and an interlaboratory adaptation does not usually lead to satisfactory test sensitivities. The aim of this study was to evaluate the influence of material variability on CXCL12 secretion by T47D cells, the outcome of the CXCL-test (estrogenic activity assay). For this purpose, the cell line sources, sera suppliers, experimental and seeding media, and the amount of cell/well were tested. The multivariable linear model (MLM), employed as an innovative approach in this field for parameter evaluation, identified that all the tested parameters had significant effects. Knowledge of the contributions of each parameter has permitted step-by-step optimization. The most beneficial approach was seeding 20,000 cells/well directly in treatment medium and using DMEM for the treatment. Great differences in both basal and maximal cytokine secretions among the three tested cell lines and different impacts of each serum were also observed. Altogether, both these biologically based and highly variable inputs were additionally assessed by MLM and a subsequent two-step evaluation, which revealed a lower variability and satisfactory reproducibility of the test. This analysis showed that not only parameter and procedure optimization but also the evaluation methodology must be considered from the perspective of interlaboratory method adaptation. This overall methodology could be applied to all bioanalytical methods for fast multiparameter and accurate analysis.
Ranked model in the form of linear transformation of multivariate feature vectors on a line can reflect a causal order between liver diseases. A priori medical knowledge about order between liver diseases and clinical data sets has been used in the definition of the convex and piecewise linear (CPL) criterion function. The linear ranked transformations have been designed here through minimization of such CPL criterion functions.
To characterize the source contributions of chemicals of emerging concern (CECs) from different aquatic environments of Taiwan, we collected water samples from different aquatic systems, which were screened for 30 pharmaceuticals and illicit drugs. The total estimated mass loadings of CECs were 23.1 g/d in southern aquatic systems and 133 g/d in central aquatic systems. We developed an analytical framework combining pollutant fingerprinting, hierarchical cluster analysis (HCA), and principal component analysis with multiple linear regression (PCA-MLR) to infer the pharmaco-signature and source contributions of CECs. Based on this approach, we estimate source contributions of 62.2% for domestic inputs, 16.9% for antibiotics application, and 20.9% for drug abuse/medication in southern aquatic system, compared with 47.3% domestic, 35.1% antibiotic, and 17.6% drug abuse/medication inputs to central aquatic systems. The proposed pharmaco-signature method provides initial insights into the profile and source apportionment of CECs in complex aquatic systems, which are of importance for environmental management.
- MeSH
- Principal Component Analysis MeSH
- Water Pollutants, Chemical analysis chemistry MeSH
- Linear Models MeSH
- Environmental Monitoring methods MeSH
- Multivariate Analysis MeSH
- Wastewater chemistry MeSH
- Rivers chemistry MeSH
- Quality Control MeSH
- Cluster Analysis MeSH
- Publication type
- Journal Article MeSH
- Research Support, Non-U.S. Gov't MeSH
- Geographicals
- Taiwan MeSH
The mixed dissociation constants of four non-steroidal anti-inflammatory drugs (NSAIDs) ibuprofen, diclofenac sodium, flurbiprofen and ketoprofen at various ionic strengths I of range 0.003-0.155, and at temperatures of 25 degrees C and 37 degrees C, were determined with the use of two different multiwavelength and multivariate treatments of spectral data, SPECFIT/32 and SQUAD(84) nonlinear regression analyses and INDICES factor analysis. The factor analysis in the INDICES program predicts the correct number of components, and even the presence of minor ones, when the data quality is high and the instrumental error is known. The thermodynamic dissociation constant pK(a)(T) was estimated by nonlinear regression of (pK(a), I) data at 25 degrees C and 37 degrees C. Goodness-of-fit tests for various regression diagnostics enabled the reliability of the parameter estimates found to be proven. PALLAS, MARVIN, SPARC, ACD/pK(a) and Pharma Algorithms predict pK(a) being based on the structural formulae of drug compounds in agreement with the experimental value. The best agreement seems to be between the ACD/pK(a) program and experimentally found values and with SPARC. PALLAS and MARVIN predicted pK(a,pred) values with larger bias errors in comparison with the experimental value for all four drugs.
- MeSH
- Anti-Inflammatory Agents, Non-Steroidal chemistry MeSH
- Models, Chemical MeSH
- Financing, Organized MeSH
- Hydrogen-Ion Concentration MeSH
- Least-Squares Analysis MeSH
- Molecular Structure MeSH
- Nonlinear Dynamics MeSH
- Solubility MeSH
- Spectrophotometry methods MeSH
- Thermodynamics MeSH
- Titrimetry methods MeSH
... revised and updated -- In the twenty years since publication of the first edition of The Statistical Analysis ... ... Models Competing Risks and Multistate Models Modeling and Analysis of Recurrent Event Data Analysis ... ... , 65 viii -- CONTENTS -- 3.6 Estimation in Log-Linear Regression Models, 68 -- 3.7 Illustrations in More ... ... Rank Regression and the Accelerated Failure Time Model 218 -- 7.1 Introduction, 218 -- 7.2 Linear Rank ... ... Analysis of Correlated Failure Time Data 302 -- 10.1 Introduction, 302 -- 10.2 Regression Models for ...
Wiley series in probability and statistics
2nd ed. xiii, 439 s.
- Keywords
- Analýza dat, Analýza statistická, Regrese,
- Conspectus
- Statistika
- NML Fields
- statistika, zdravotnická statistika
