Bayesian inference for multivariate meta‐analysis Box–Cox transformation models for individual patient data with applications to evaluation of cholesterol‐lowering drugs

Kim, Sungduk; Chen, Ming Hui; Ibrahim, Joseph G.; Shah, Arvind; Lin, Jianxin

doi:10.1002/sim.5814

Cited by 11 publications

(30 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other. Several one-stage models have been proposed for IPD-MA where the outcome belongs to another type of non-normally distributed data , recurrent events (Haines and Hill, 2011), repeated measurements , or multivariate responses (Kim et al, 2013;Riley et al, 2008;Yamaguchi et al, 2002).…”

Section: Countmentioning

confidence: 99%

Get real in individual participant data (IPD) meta‐analysis: a review of the methodology

Debray

Moons

Valkenhoef

et al. 2015

Research Synthesis Methods

278

326

View full text Add to dashboard Cite

Individual participant data (IPD) meta‐analysis is an increasingly used approach for synthesizing and investigating treatment effect estimates. Over the past few years, numerous methods for conducting an IPD meta‐analysis (IPD‐MA) have been proposed, often making different assumptions and modeling choices while addressing a similar research question. We conducted a literature review to provide an overview of methods for performing an IPD‐MA using evidence from clinical trials or non‐randomized studies when investigating treatment efficacy. With this review, we aim to assist researchers in choosing the appropriate methods and provide recommendations on their implementation when planning and conducting an IPD‐MA. © 2015 The Authors. Research Synthesis Methods published by John Wiley & Sons, Ltd.

show abstract

Section: Countmentioning

confidence: 99%

Get real in individual participant data (IPD) meta‐analysis: a review of the methodology

Debray

Moons

Valkenhoef

et al. 2015

Research Synthesis Methods

278

326

View full text Add to dashboard Cite

show abstract

“…These L measures indicate that (i) the models with an unstructured Σ fit the cholesterol data much better than the respective models with a diagonal Σ; (ii) under an unstructured Σ, the t model fit the data better than the normal model; (iii) under both the diagonal Σ and unstructured Σ, the skew t models fit the data better than the symmetric normal and t models; (iv) under both the diagonal Σ and unstructured Σ, the skew t models with dependent z ijk ’s ( τ < ∞) outperformed the corresponding models with independent z ijk ’s ( τ = ∞ ); and (v) among the 9 models with a diagonal Σ, the skew t model with τ = 25 was the best while the skew t model with τ = 30 was the best among the 9 models with an unstructured Σ according to the L measure. These results are quite interesting and appealing since they showed that the three outcome variables were indeed dependent and the distributions of these variables were not symmetric for the cholesterol data as shown in Kim et al 1…”

Section: Analysis Of the Cholesterol Datamentioning

confidence: 70%

“…The proposed model includes multivariate normal meta-regression models, multivariate skew normal meta-regression models; and multivariate skew t meta-regression models as special cases. Compared to Kim et al 1 who used a Box-Cox transformation on the response variables, our proposed multivariate skew t meta-regression model is more attractive since the point estimates are more directly comparable and easier to interpret than those under a Box-Cox transformation model and the proposed model requires much less computational time. Estimation of the parameters requires a sophisticated and computationally intensive Markov chain Monte Carlo (MCMC) sampling algorithm.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Bayesian multivariate skew meta-regression models for individual patient data

Ibrahim

Kim

Chen

et al. 2018

Stat Methods Med Res

Self Cite

View full text Add to dashboard Cite

We examine a class of multivariate meta-regression models in the presence of Individual Patient Data (IPD). The methodology is well motivated from several studies of cholesterol lowering drugs where the goal is to jointly analyze the multivariate outcomes, Low Density Lipoprotein Cholesterol (LDL-C), High Density Lipoprotein Cholesterol (HDL-C), and Triglycerides (TG). These three continuous outcome measures are correlated and shed much light on a subject’s lipid status. One of the main goals in lipid research is the joint analysis of these three outcome measures in a meta-regression setting. Since these outcome measures are not typically multivariate normal, one must consider classes of distributions that allow for skewness in one or more of the outcomes. In this paper, we consider a new general class of multivariate skew distributions for multivariate meta-regression and examine their theoretical properties. Using these distributions, we construct a Bayesian model for the meta-data and develop an efficient Markov chain Monte Carlo (MCMC) computational scheme for carrying out the computations. In addition, we develop a multivariate L measure for model comparison, Bayesian residuals for model assessment, and a Bayesian procedure for detecting outlying trials. The proposed multivariate L measure, Bayesian residuals, and Bayesian outlying trial detection procedure are particularly suitable and computationally attractive in the multivariate meta-regression setting. A detailed case study demonstrating the usefulness of the proposed methodology is carried out in an IPD multivariate meta-regression setting using 26 pivotal Merck clinical trials that compare statins (cholesterol lowering drugs) in combination with ezetimibe and statins alone on treatment-naïve patients and those continuing on statins at baseline.

show abstract

“…We first performed a Box-Cox Power Transformation [17] using a powerTransform function (car package in R 3.4.0) to make the distribution of each mRNA and lncRNA in each sample approximately normal.…”

Section: Methodsmentioning

confidence: 99%

Identification of long non-coding RNA-related and –coexpressed mRNA biomarkers for hepatocellular carcinoma

et al. 2019

View full text Add to dashboard Cite

BackgroundWhile changes in mRNA expression during tumorigenesis have been used widely as molecular biomarkers for the diagnosis of a number of cancers, the approach has limitations. For example, traditional methods do not consider the regulatory and positional relationship between mRNA and lncRNA. The latter has been largely shown to possess tumor suppressive or oncogenic properties. The combined analysis of mRNA and lncRNA is likely to facilitate the identification of biomarkers with higher confidence.ResultsTherefore, we have developed an lncRNA-related method to identify traditional mRNA biomarkers. First we identified mRNAs that are differentially expressed in Hepatocellular Carcinoma (HCC) by comparing cancer and matched adjacent non-tumorous liver tissues. Then, we performed mRNA-lncRNA relationship and coexpression analysis and obtained 41 lncRNA-related and -coexpressed mRNA biomarkers. Next, we performed network analysis, gene ontology analysis and pathway analysis to unravel the functional roles and molecular mechanisms of these lncRNA-related and -coexpressed mRNA biomarkers. Finally, we validated the prediction and performance of the 41 lncRNA-related and -coexpressed mRNA biomarkers using Support Vector Machine model with five-fold cross-validation in an independent HCC dataset from RNA-seq.ConclusionsOur results suggested that mRNAs expression profiles coexpressed with positionally related lncRNAs can provide important insights into early diagnosis and specific targeted gene therapy of HCC.

show abstract

Bayesian inference for multivariate meta‐analysis Box–Cox transformation models for individual patient data with applications to evaluation of cholesterol‐lowering drugs

Cited by 11 publications

References 24 publications

Get real in individual participant data (IPD) meta‐analysis: a review of the methodology

Get real in individual participant data (IPD) meta‐analysis: a review of the methodology

Bayesian multivariate skew meta-regression models for individual patient data

Identification of long non-coding RNA-related and –coexpressed mRNA biomarkers for hepatocellular carcinoma

Contact Info

Product

Resources

About