Jacobs P, Viechtbauer W. Estimation of the biserial correlation and its sampling variance for use in meta-analysis. Res.Synth.Method. Epub 2016 Sep 15. PMID: 27631635.

Meta-analyses are often used to synthesize the findings of studies examining the correlational relationship between two continuous variables. When only dichotomous measurements are available for one of the two variables, the biserial correlation coefficient can be used to estimate the product-moment correlation between the two underlying continuous variables. Unlike the point-biserial correlation coefficient, biserial correlation coefficients can therefore be integrated with product-moment correlation coefficients in the same meta-analysis. The present article describes the estimation of the biserial correlation coefficient for meta-analytic purposes and reports simulation results comparing different methods for estimating the coefficient's sampling variance. The findings indicate that commonly employed methods yield inconsistent estimates of the sampling variance across a broad range of research situations. In contrast, consistent estimates can be obtained using two methods that appear to be unknown in the meta-analytic literature. A variance-stabilizing transformation for the biserial correlation coefficient is described that allows for the construction of confidence intervals for individual coefficients with close to nominal coverage probabilities in most of the examined conditions. Copyright (c) 2016 John Wiley & Sons, Ltd.

DOI: https://dx.doi.org/10.1002/jrsm.1218.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27631635.

Friede T, Röver C, Wandel S, Neuenschwander B. Meta-analysis of two studies in the presence of heterogeneity with applications in rare diseases. Biom.J. Epub 2016 Oct 18. PMID: 27754556.

Random-effects meta-analyses are used to combine evidence of treatment effects from multiple studies. Since treatment effects may vary across trials due to differences in study characteristics, heterogeneity in treatment effects between studies must be accounted for to achieve valid inference. The standard model for random-effects meta-analysis assumes approximately normal effect estimates and a normal random-effects model. However, standard methods based on this model ignore the uncertainty in estimating the between-trial heterogeneity. In the special setting of only two studies and in the presence of heterogeneity, we investigate here alternatives such as the Hartung-Knapp-Sidik-Jonkman method (HKSJ), the modified Knapp-Hartung method (mKH, a variation of the HKSJ method) and Bayesian random-effects meta-analyses with priors covering plausible heterogeneity values; R code to reproduce the examples is presented in an appendix. The properties of these methods are assessed by applying them to five examples from various rare diseases and by a simulation study. Whereas the standard method based on normal quantiles has poor coverage, the HKSJ and mKH generally lead to very long, and therefore inconclusive, confidence intervals. The Bayesian intervals on the whole show satisfying properties and offer a reasonable compromise between these two extremes.

DOI: https://doi.org/10.1002/bimj.201500236.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27754556.

Shrier I, Christensen R, Juhl C, Beyene J. Meta-analysis on continuous outcomes in minimal important difference units: An application with appropriate variance calculations. J.Clin.Epidemiol. Epub 2016 Jul 29. PMID: 27480962.

To compare results from meta-analyses for mean differences in minimal important difference (MID) units (MDMID), when MID is treated as a random variable versus a constant.
Meta-analyses of published data. We calculated the variance of MDMID as a random variable using the delta method, and as a constant. We assessed performance under different assumptions. We compare meta-analysis results from data originally used to present the MDMID, and data from osteoarthritis studies using different domain instruments.
Depending on the data set and depending on the values of rho and CoV-MID, estimates of treatment effect and p-values between an approach considering the MID as a constant versus as a random variable may differ appreciably. Using our data sets, we provide examples of the potential magnitude. When rho=0.5 and CoVMID=0.8, considering MID as a constant overestimated the treatment effect by 33%-110%, and decreased the p-value for heterogeneity from above 0.95 to below 0.08. When rho=0.8 and CoVMID=0.5, the magnitude of the effects were similar.
Considering MID as a random variable avoids unrealistic assumptions and provides more appropriate treatment effect estimates.
Copyright © 2016 Elsevier Inc. All rights reserved.

DOI: http://dx.doi.org/10.1016/j.jclinepi.2016.07.012.
PubMed: http://www.ncbi.nlm.nih.gov/pubmed/27480962.

Dogo SH, Clark A, Kulinskaya E. Sequential change detection and monitoring of temporal trends in random-effects meta-analysis. Res.Synth.Methods. Epub 2016 Dec 8. PMID: 27933728.

Temporal changes in magnitude of effect sizes reported in many areas of research are a threat to the credibility of the results and conclusions of meta-analysis. Numerous sequential methods for meta-analysis have been proposed to detect changes and monitor trends in effect sizes so that meta-analysis can be updated when necessary and interpreted based on the time it was conducted. The difficulties of sequential meta-analysis under the random-effects model are caused by dependencies in increments introduced by the estimation of the heterogeneity parameter tau2 . In this paper, we propose the use of a retrospective cumulative sum (CUSUM)-type test with bootstrap critical values. This method allows retrospective analysis of the past trajectory of cumulative effects in random-effects meta-analysis and its visualization on a chart similar to CUSUM chart. Simulation results show that the new method demonstrates good control of Type I error regardless of the number or size of the studies and the amount of heterogeneity. Application of the new method is illustrated on two examples of medical meta-analyses. (c) 2016 The Authors. Research Synthesis Methods published by John Wiley & Sons Ltd.

DOI: https://dx.doi.org/10.1002/jrsm.1222.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27933728.

Yoneoka D, Henmi M. Synthesis of linear regression coefficients by recovering the within-study covariance matrix from summary statistics. Res.Synth.Methods. Epub 2016 Dec 16. PMID: 27987264.

Recently, the number of regression models has dramatically increased in several academic fields. However, within the context of meta-analysis, synthesis methods for such models have not been developed in a commensurate trend. One of the difficulties hindering the development is the disparity in sets of covariates among literature models. If the sets of covariates differ across models, interpretation of coefficients will differ, thereby making it difficult to synthesize them. Moreover, previous synthesis methods for regression models, such as multivariate meta-analysis, often have problems because covariance matrix of coefficients (i.e. within-study correlations) or individual patient data are not necessarily available. This study, therefore, proposes a brief explanation regarding a method to synthesize linear regression models under different covariate sets by using a generalized least squares method involving bias correction terms. Especially, we also propose an approach to recover (at most) threecorrelations of covariates, which is required for the calculation of the bias term without individual patient data. Copyright (c) 2016 John Wiley & Sons, Ltd.

DOI: https://dx.doi.org/10.1002/jrsm.1228.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27987264.

Borenstein M, Higgins JP, Hedges LV, Rothstein HR. Basics of meta-analysis: I2 is not an absolute measure of heterogeneity. Res.Synth.Methods. 2017 Mar;8(1):5-18. PMID: 28058794.

When we speak about heterogeneity in a meta-analysis, our intent is usually to understand the substantive implications of the heterogeneity. If an intervention yields a mean effect size of 50 points, we want to know if the effect size in different populations varies from 40 to 60, or from 10 to 90, because this speaks to the potential utility of the intervention. While there is a common belief that the I2 statistic provides this information, it actually does not. In this example, if we are told that I2 is 50%, we have no way of knowing if the effects range from 40 to 60, or from 10 to 90, or across some other range. Rather, if we want to communicate the predicted range of effects, then we should simply report this range. This gives readers the information they think is being captured by I2 and does so in a way that is concise and unambiguous. Copyright © 2017 John Wiley & Sons, Ltd.
Copyright © 2017 John Wiley & Sons, Ltd

DOI: https://dx.doi.org/10.1002/jrsm.1230.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28058794.

Choi SW, Lam DM. Heterogeneity in meta-analyses. Comparing apples and oranges? Anaesthesia. 2017 Apr;72(4):532-4. PMID: 28213890.
DOI: http://dx.doi.org/10.1111/anae.13832.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=28213890.

Feingold A. Meta-Analysis With Standardized Effect Sizes From Multilevel and Latent Growth Models. J.Consult.Clin.Psychol. 2017 Mar;85(3):262-6. PMID: 28068118.

Findings from multilevel and latent growth modeling analysis (GMA) need to be included in literature reviews, and this article explicates 4 rarely discussed approaches for using GMA studies in meta-analysis.
Extant and new equations are presented for calculating the effect size (d) and its variance (v) from reported statistics from GMA studies with each method, and a fixed effects meta-analysis of results from 5 randomized clinical trials was conducted to demonstrate their applications.
Two common problematic practices--one that introduces bias in effect sizes because of attrition, measurement errors, and probable violations of assumptions for classical analysis, and the other that confounds the treatment effect with the intraclass correlation--were both found to yield smaller effect sizes from retrieved studies than were obtained with a newer model-based framework and its associated GMA d statistic.
The optimal strategy for including a GMA study in a meta-analysis is to use GMA d and its v calculated with the standard error of the unstandardized coefficient for the treatment effect. When that standard error is unknown, the use of GMA d and its v estimated with an alternative equation that requires only GMA d and sample size is recommended. (PsycINFO Database Record
(c) 2017 APA, all rights reserved).

DOI: https://dx.doi.org/10.1037/ccp0000162.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=28068118.

Friede T, Röver C, Wandel S, Neuenschwander B. Meta-analysis of few small studies in orphan diseases. Res.Synth.Method. 2017 Mar;8(1):79-91. PMID: 27362487.

Meta-analyses in orphan diseases and small populations generally face particular problems, including small numbers of studies, small study sizes and heterogeneity of results. However, the heterogeneity is difficult to estimate if only very few studies are included. Motivated by a systematic review in immunosuppression following liver transplantation in children, we investigate the properties of a range of commonly used frequentist and Bayesian procedures in simulation studies. Furthermore, the consequences for interval estimation of the common treatment effect in random-effects meta-analysis are assessed. The Bayesian credibility intervals using weakly informative priors for the between-trial heterogeneity exhibited coverage probabilities in excess of the nominal level for a range of scenarios considered. However, they tended to be shorter than those obtained by the Knapp-Hartung method, which were also conservative. In contrast, methods based on normal quantiles exhibited coverages well below the nominal levels in many scenarios. With very few studies, the performance of the Bayesian credibility intervals is of course sensitive to the specification of the prior for the between-trial heterogeneity. In conclusion, the use of weakly informative priors as exemplified by half-normal priors (with a scale of 0.5 or 1.0) for log odds ratios is recommended for applications in rare diseases. (c) 2016 The Authors. Research Synthesis Methods published by John Wiley & Sons Ltd.

DOI: http://dx.doi.org/10.1002/jrsm.1217.
PubMed: http://www.ncbi.nlm.nih.gov/pubmed/?term=27362487.

Partlett C, Riley RD. Random effects meta-analysis: Coverage performance of 95% confidence and prediction intervals following REML estimation. Stat.Med. 2017 Jan 30;36(2):301-17. PMID: 27714841.

A random effects meta-analysis combines the results of several independent studies to summarise the evidence about a particular measure of interest, such as a treatment effect. The approach allows for unexplained between-study heterogeneity in the true treatment effect by incorporating random study effects about the overall mean. The variance of the mean effect estimate is conventionally calculated by assuming that the between study variance is known; however, it has been demonstrated that this approach may be inappropriate, especially when there are few studies. Alternative methods that aim to account for this uncertainty, such as Hartung-Knapp, Sidik-Jonkman and Kenward-Roger, have been proposed and shown to improve upon the conventional approach in some situations. In this paper, we use a simulation study to examine the performance of several of these methods in terms of the coverage of the 95% confidence and prediction intervals derived from a random effects meta-analysis estimated using restricted maximum likelihood. We show that, in terms of the confidence intervals, the Hartung-Knapp correction performs well across a wide-range of scenarios and outperforms other methods when heterogeneity was large and/or study sizes were similar. However, the coverage of the Hartung-Knapp method is slightly too low when the heterogeneity is low (I2 30%) and study sizes are similar. In other situations, especially when heterogeneity is small and the study sizes are quite varied, the coverage is far too low and could not be consistently improved by either increasing the number of studies, altering the degrees of freedom or using variance inflation methods. Therefore, researchers should be cautious in deriving 95% prediction intervals following a frequentist random-effects meta-analysis until a more reliable solution is identified. (c) 2016 The Authors. Statistics in Medicine Published by John Wiley & Sons Ltd.

FREE FULL TEXT: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5157768/pdf/SIM-36-301.pdf
DOI: https://doi.org/10.1002/sim.7140.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27714841.
PubMed Central: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5157768/.

Walwyn R, Roberts C. Meta-analysis of standardised mean differences from randomised trials with treatment-related clustering associated with care providers. Stat.Med. 2017 Mar 30;36(7):1043-67. PMID: 27910117.

In meta-analyses, where a continuous outcome is measured with different scales or standards, the summary statistic is the mean difference standardised to a common metric with a common variance. Where trial treatment is delivered by a person, nesting of patients within care providers leads to clustering that may interact with, or be limited to, one or more of the arms. Assuming a common standardising variance is less tenable and options for scaling the mean difference become numerous. Metrics suggested for cluster-randomised trials are within, between and total variances and for unequal variances, the control arm or pooled variances. We consider summary measures and individual-patient-data methods for meta-analysing standardised mean differences from trials with two-level nested clustering, relaxing independence and common variance assumptions, allowing sample sizes to differ across arms. A general metric is proposed with comparable interpretation across designs. The relationship between the method of standardisation and choice of model is explored, allowing for bias in the estimator and imprecision in the standardising metric. A meta-analysis of trials of counselling in primary care motivated this work. Assuming equal clustering effects across trials, the proposed random-effects meta-analysis model gave a pooled standardised mean difference of -0.27 (95% CI -0.45 to -0.08) using summary measures and -0.26 (95% CI -0.45 to -0.09) with the individual-patient-data. While treatment-related clustering has rarely been taken into account in trials, it is now recommended that it is considered in trials and meta-analyses. This paper contributes to the uptake of this guidance. Copyright (c) 2016 John Wiley & Sons, Ltd.

DOI: https://doi.org/10.1002/sim.7186.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27910117.

Baurecht H, Hotze M, Rodríguez E, Manz J, Weidinger S, Cordell HJ, Augustin T, Strauch K. Compare and Contrast Meta Analysis (CCMA): A Method for Identification of Pleiotropic Loci in Genome-Wide Association Studies. PLoS One. 2016 May 5;11(5):e0154872. PMID: 27149374.

In recent years, genome-wide association studies (GWAS) have identified many loci that are shared among common disorders and this has raised interest in pleiotropy. For performing appropriate analysis, several methods have been proposed, e.g. conducting a look-up in external sources or exploiting GWAS results by meta-analysis based methods. We recently proposed the Compare & Contrast Meta-Analysis (CCMA) approach where significance thresholds were obtained by simulation. Here we present analytical formulae for the density and cumulative distribution function of the CCMA test statistic under the null hypothesis of no pleiotropy and no association, which, conveniently for practical reasons, turns out to be exponentially distributed. This allows researchers to apply the CCMA method without having to rely on simulations. Finally, we show that CCMA demonstrates power to detect disease-specific, agonistic and antagonistic loci comparable to the frequently used Subset-Based Meta-Analysis approach, while better controlling the type I error rate.

FREE FULL TEXT: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4858294/pdf/pone.0154872.pdf
DOI: https://doi.org/10.1371/journal.pone.0154872.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27149374.
PubMed Central: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4858294/.

Cam Pham N, Haibe-Kains B, Bellot P, Bontempi G, Meyer PE. Study of Meta-analysis Strategies for Network Inference Using Information-Theoretic Approaches. Study of Meta-analysis Strategies for Network Inference Using Information-Theoretic Approaches. Proceedings of the 27th International Workshop on Database and Expert Systems Applications (DEXA); 2016 Sep 5-8; Porto, Portugal. [New York, NY]: IEEE;2016 :76-83.

Reverse engineering of gene regulatory networks (GRNs) from gene expression data is a classical challenge in systems biology. Thanks to high-throughput technologies, a massive amount of gene-expression data has been accumulated in the public repositories. Modelling GRNs from multiple experiments (also called integrative analysis) has, therefore, naturally become a standard procedure in modern computational biology. Indeed, such analysis is usually more robust than the traditional approaches focused on individual datasets, which typically suffer from some experimental bias and a small number of samples. To date, there are mainly two strategies for the problem of interest: the first one ("data merging") merges all data sets together and then infers a GRN whereas the other ("networks ensemble") infers GRNs from every dataset separately and then aggregates them using some ensemble rules (such as ranksumor weightsum). Unfortunately, a thorough comparison of these two approaches is lacking. In this paper, we evaluate the performances of various meta-analysis approaches mentioned above with a systematic set of experiments based on in silico benchmarks. Furthermore, we present a new meta-analysis approach for inferring GRNs from multiple studies. Our proposed approach, adapted to methods based on pairwise measures such as correlation or mutual information, consists of two steps: aggregating matrices of the pairwise measures from every dataset followed by extracting the network from the meta-matrix.

INSPEC Accession Number: 16600553
DOI: https://doi.org/10.1109/DEXA.2016.030.

Cho H, Kim H, Na D, Kim SY, Jo D, Lee D. Meta-analysis method for discovering reliable biomarkers by integrating statistical and biological approaches: An application to liver toxicity. Biochem.Biophys.Res.Commun. 2016 Mar 4;471(2):274-81. PMID: 26820531.

Biomarkers that are identified from a single study often appear to be biologically irrelevant or false positives. Meta-analysis techniques allow integrating data from multiple studies that are related but independent in order to identify biomarkers across multiple conditions. However, existing biomarker meta-analysis methods tend to be sensitive to the dataset being analyzed. Here, we propose a meta-analysis method, iMeta, which integrates t-statistic and fold change ratio for improved robustness. For evaluation of predictive performance of the biomarkers identified by iMeta, we compare our method with other meta-analysis methods. As a result, iMeta outperforms the other methods in terms of sensitivity and specificity, and especially shows robustness to study variance increase; it consistently shows higher classification accuracy on diverse datasets, while the performance of the others is highly affected by the dataset being analyzed. Application of iMeta to 59 drug-induced liver injury studies identified three key biomarker genes: Zwint, Abcc3, and Ppp1r3b. Experimental evaluation using RT-PCR and qRT-PCR shows that their expressional changes in response to drug toxicity are concordant with the result of our method. iMeta is available at http://imeta.kaist.ac.kr/index.html.

FREE FULL TEXT: http://ac.els-cdn.com/S0006291X16300821/1-s2.0-S0006291X16300821-main.pdf?_tid=9da13b26-55f8-11e7-afef-00000aacb362&acdnat=1497991374_44f8cc16ae02be260a6d92e70fc931cb
DOI: https://dx.doi.org/10.1016/j.bbrc.2016.01.082.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=26820531.

Fu R, Holmer HK. Change score or follow-up score? Choice of mean difference estimates could impact meta-analysis conclusions. J.Clin.Epidemiol. 2016 Aug;76:108-17. PMID: 26931293.

In randomized controlled clinical trials, continuous outcomes are typically measured at both baseline and follow-up, and mean difference could be estimated using the change scores from baseline or the follow-up scores. This study assesses the impact of using change score vs. follow-up score on the conclusions of meta-analyses.
A total of 63 meta-analyses from six comparative effectiveness reviews were included. The combined mean difference was estimated using a random-effects model, and we also evaluated whether the impact qualitatively varied by alternative random-effects estimates.
Based on the Dersimonian-Laird (DL) method, using the change vs. the follow-up score led to five meta-analyses (7.9%) showing discrepancy in conclusions. Based on the profile likelihood (PL) method, nine (14.3%) showed discrepancy in conclusions. Using change score was more likely to show a significant difference in effects between interventions (DL method: 4 of 5; PL method: 7 of 9). A significant difference in baseline scores did not necessarily lead to discrepancies in conclusions.
Using the change vs. the follow-up score could lead to important discrepancies in conclusions. Sensitivity analyses should be conducted to check the robustness of results to the choice of mean difference estimates.
Copyright © 2016 Elsevier Inc. All rights reserved.

DOI: https://dx.doi.org/10.1016/j.jclinepi.2016.01.034.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26931293.

George BJ, Aban IB. An application of meta-analysis based on DerSimonian and Laird method. J.Nucl.Cardiol. 2016 Aug;23(4):690-2. PMID: 26245193.
DOI: https://doi.org/10.1007/s12350-015-0249-6.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/26245193.

Jackson D, Bowden J. Confidence intervals for the between-study variance in random-effects meta-analysis using generalised heterogeneity statistics: should we use unequal tails? BMC Med.Res.Methodol. 2016 Sep 7;16(1):118. PMID: 27604952.

Confidence intervals for the between study variance are useful in random-effects meta-analyses because they quantify the uncertainty in the corresponding point estimates. Methods for calculating these confidence intervals have been developed that are based on inverting hypothesis tests using generalised heterogeneity statistics. Whilst, under the random effects model, these new methods furnish confidence intervals with the correct coverage, the resulting intervals are usually very wide, making them uninformative.
We discuss a simple strategy for obtaining 95 % confidence intervals for the between-study variance with a markedly reduced width, whilst retaining the nominal coverage probability. Specifically, we consider the possibility of using methods based on generalised heterogeneity statistics with unequal tail probabilities, where the tail probability used to compute the upper bound is greater than 2.5 %. This idea is assessed using four real examples and a variety of simulation studies. Supporting analytical results are also obtained.
Our results provide evidence that using unequal tail probabilities can result in shorter 95 % confidence intervals for the between-study variance. We also show some further results for a real example that illustrates how shorter confidence intervals for the between-study variance can be useful when performing sensitivity analyses for the average effect, which is usually the parameter of primary interest.
We conclude that using unequal tail probabilities when computing 95 % confidence intervals for the between-study variance, when using methods based on generalised heterogeneity statistics, can result in shorter confidence intervals. We suggest that those who find the case for using unequal tail probabilities convincing should use the '1-4 % split', where greater tail probability is allocated to the upper confidence bound. The 'width-optimal' interval that we present deserves further investigation.

FREE FULL TEXT: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5015418/pdf/12874_2016_Article_219.pdf
DOI: https://doi.org/10.1186/s12874-016-0219-y.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27604952.
PubMed Central: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5015418/.

Kim S, Lin CW, Tseng GC. MetaKTSP: a meta-analytic top scoring pair method for robust cross-study validation of omics prediction analysis. Bioinformatics. 2016 Jul 1;32(13):1966-73. PMID: 27153719.

Supervised machine learning is widely applied to transcriptomic data to predict disease diagnosis, prognosis or survival. Robust and interpretable classifiers with high accuracy are usually favored for their clinical and translational potential. The top scoring pair (TSP) algorithm is an example that applies a simple rank-based algorithm to identify rank-altered gene pairs for classifier construction. Although many classification methods perform well in cross-validation of single expression profile, the performance usually greatly reduces in cross-study validation (i.e. the prediction model is established in the training study and applied to an independent test study) for all machine learning methods, including TSP. The failure of cross-study validation has largely diminished the potential translational and clinical values of the models. The purpose of this article is to develop a meta-analytic top scoring pair (MetaKTSP) framework that combines multiple transcriptomic studies and generates a robust prediction model applicable to independent test studies.
We proposed two frameworks, by averaging TSP scores or by combining P-values from individual studies, to select the top gene pairs for model construction. We applied the proposed methods in simulated data sets and three large-scale real applications in breast cancer, idiopathic pulmonary fibrosis and pan-cancer methylation. The result showed superior performance of cross-study validation accuracy and biomarker selection for the new meta-analytic framework. In conclusion, combining multiple omics data sets in the public domain increases robustness and accuracy of the classification model that will ultimately improve disease understanding and clinical treatment decisions to benefit patients.
An R package MetaKTSP is available online. (http://tsenglab.biostat.pitt.edu/software.htm).
Supplementary data are available at Bioinformatics online.
© The Author 2016. Published by Oxford University Press. All rights reserved. For Permissions, please e-mail: journals.permissions@oup.com.

DOI: https://doi.org/10.1093/bioinformatics/btw115.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27153719.

Lagani V, Karozou AD, Gomez-Cabrero D, Silberberg G, Tsamardinos I. A comparative evaluation of data-merging and meta-analysis methods for reconstructing gene-gene interactions. BMC Bioinformatics. 2016 Jun 6;17(Suppl 5):194. PMID: 27294826.

We address the problem of integratively analyzing multiple gene expression, microarray datasets in order to reconstruct gene-gene interaction networks. Integrating multiple datasets is generally believed to provide increased statistical power and to lead to a better characterization of the system under study. However, the presence of systematic variation across different studies makes network reverse-engineering tasks particularly challenging. We contrast two approaches that have been frequently used in the literature for addressing systematic biases: meta-analysis methods, which first calculate opportune statistics on single datasets and successively summarize them, and data-merging methods, which directly analyze the pooled data after removing eventual biases. This comparative evaluation is performed on both synthetic and real data, the latter consisting of two manually curated microarray compendia comprising several E. coli and Yeast studies, respectively. Furthermore, the reconstruction of the regulatory network of the transcription factor Ikaros in human Peripheral Blood Mononuclear Cells (PBMCs) is presented as a case-study.
The meta-analysis and data-merging methods included in our experimentations provided comparable performances on both synthetic and real data. Furthermore, both approaches outperformed (a) the naïve solution of merging data together ignoring possible biases, and (b) the results that are expected when only one dataset out of the available ones is analyzed in isolation. Using correlation statistics proved to be more effective than using p-values for correctly ranking candidate interactions. The results from the PBMC case-study indicate that the findings of the present study generalize to different types of network reconstruction algorithms.
Ignoring the systematic variations that differentiate heterogeneous studies can produce results that are statistically indistinguishable from random guessing. Meta-analysis and data merging methods have proved equally effective in addressing this issue, and thus researchers may safely select the approach that best suit their specific application.

Erratum in:

Lagani V, Karozou AD, Gomez-Cabrero D, Silberberg G, Tsamardinos I. Erratum to: A comparative evaluation of data-merging and meta-analysis methods for reconstructing gene-gene interactions. BMC Bioinformatics. 2016 Jul 27;17:290. doi: 10.1186/s12859-016-1153-z. PubMed PMID: 27465624; PubMed Central PMCID: PMC4963931.

FREE FULL TEXT: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4905611/pdf/12859_2016_Article_1038.pdf
DOI: https://doi.org/10.1186/s12859-016-1038-1.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27294826.
PubMed Central: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4905611/.

Liu Y, Chen Y, Scheet P. A meta-analytic framework for detection of genetic interactions. Genet.Epidemiol. 2016 Nov;40(7):534-43. PMID: 27528046.

With varying, but substantial, proportions of heritability remaining unexplained by summaries of single-SNP genetic variation, there is a demand for methods that extract maximal information from genetic association studies. One source of variation that is difficult to assess is genetic interactions. A major challenge for naive detection methods is the large number of possible combinations, with a requisite need to correct for multiple testing. Assumptions of large marginal effects, to reduce the search space, may be restrictive and miss higher order interactions with modest marginal effects. In this paper, we propose a new procedure for detecting gene-by-gene interactions through heterogeneity in estimated low-order (e.g., marginal) effect sizes by leveraging population structure, or ancestral differences, among studies in which the same phenotypes were measured. We implement this approach in a meta-analytic framework, which offers numerous advantages, such as robustness and computational efficiency, and is necessary when data-sharing limitations restrict joint analysis. We effectively apply a dimension reduction procedure that scales to allow searches for higher order interactions. For comparison to our method, which we term phylogenY-aware Effect-size Tests for Interactions (YETI), we adapt an existing method that assumes interacting loci will exhibit strong marginal effects to our meta-analytic framework. As expected, YETI excels when multiple studies are from highly differentiated populations and maintains its superiority in these conditions even when marginal effects are small. When these conditions are less extreme, the advantage of our method wanes. We assess the Type-I error and power characteristics of complementary approaches to evaluate their strengths and limitations.

DOI: http://dx.doi.org/10.1002/gepi.21996.
PubMed: http://www.ncbi.nlm.nih.gov/pubmed/27528046.

Morrissey MB. Meta-analysis of magnitudes, differences and variation in evolutionary parameters. J.Evol.Biol. 2016 Oct;29(10):1882-904. PMID: 27726237.

Meta-analysis is increasingly used to synthesize major patterns in the large literatures within ecology and evolution. Meta-analytic methods that do not account for the process of observing data, which we may refer to as 'informal meta-analyses', may have undesirable properties. In some cases, informal meta-analyses may produce results that are unbiased, but do not necessarily make the best possible use of available data. In other cases, unbiased statistical noise in individual reports in the literature can potentially be converted into severe systematic biases in informal meta-analyses. I first present a general description of how failure to account for noise in individual inferences should be expected to lead to biases in some kinds of meta-analysis. In particular, informal meta-analyses of quantities that reflect the dispersion of parameters in nature, for example, the mean absolute value of a quantity, are likely to be generally highly misleading. I then re-analyse three previously published informal meta-analyses, where key inferences were of aspects of the dispersion of values in nature, for example, the mean absolute value of selection gradients. Major biological conclusions in each original informal meta-analysis closely match those that could arise as artefacts due to statistical noise. I present alternative mixed-model-based analyses that are specifically tailored to each situation, but where all analyses may be implemented with widely available open-source software. In each example meta-re-analysis, major conclusions change substantially.

DOI: https://dx.doi.org/10.1111/jeb.12950.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27726237.

Nakagawa S, Lagisz M. Visualizing unbiased and biased unweighted meta-analyses. J.Evol.Biol. 2016 Oct;29(10):1914-6. PMID: 27397701.

[First paragraph]

Most meta-analyses in ecology and evolution are concerned with an overall effect (average effect sizes) of a biological phenomenon. However, some meta-analyses are interested in average magnitudes or the mean of absolute effect sizes. For example, in a meta-analysis of selection gradients where signs of such gradients can be arbitrary, one uses absolute values of selection gradients as effect size (e.g. Kingsolver et al., 2001). Michael Morrissey’s Target Review (2016) clearly demonstrates that when average magnitudes are of interest, one must correct for the statistical noise (sampling error) of effect sizes. Otherwise, meta-analytic results, that is average magnitudes, will be biased. This is in contrast to a classical situation of overall average of effect sizes, which is unbiased when not accounting for sampling error (variance). Morrissey terms meta-analyses that do not deal with sampling error as ‘informal meta-analysis’. Such meta-analysis is also referred to as unweighted meta-analysis, as effect sizes are not weighted by the inverse of corresponding sampling error variances (e.g. Jennions et al., 2001).

DOI: https://doi.org/10.1111/jeb.12945.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=27397701.

Nguyen T, Diaz D, Draghici S. TOMAS: A novel topology-aware meta-analysis approach applied to system biology. TOMAS: A novel topology-aware meta-analysis approach applied to system biology. Proceedings of the BCB '16: Proceedings of the 7th ACM International Conference on Bioinformatics, Computational Biology, and Health Informatics; 2016 October 2-5; Seattle, Washington. New York, NY: ACM;2016 :13-22.

With the explosion of high-throughput data, an effective integrative analysis is needed to decipher the knowledge accumulated in multiple studies. However, batch effects, patient heterogeneity, and disease complexity all complicate the integration of data from different sources. Here we introduce TOMAS, a novel meta-analysis framework that transforms the challenging meta-analysis problem into a set of standard analysis problems that can be solved efficiently. This framework utilizes techniques based on both p-values and effect sizes to identify differentially expressed genes and their expression change on a genome-scale. The computed statistics allow for topology-aware pathway analysis of the given phenotypes, where topological information of genes is taken into consideration. We compare TOMAS with four meta-analysis approaches, as well as with three dedicated pathway analysis approaches that employ multiple datasets (MetaPath). The eight approaches have been tested on 609 samples from 9 Alzheimer's studies conducted in independent labs for different sets of patients and tissues. We demonstrate that the topology based meta-analysis framework overcomes noise and bias to identify pathways that are known to be implicated in Alzheimer's disease. While presented here in a genomic data analysis application, the proposed framework is sufficiently general to be applied in other research areas.

DOI: https://dx.doi.org/10.1145/2975167.2975168.

Ray D,Boehnke M. A Unified Association test for the Meta-Analysis of Multiple Traits using GWAS Summary Statistics. A Unified Association test for the Meta-Analysis of Multiple Traits using GWAS Summary Statistics. Proceedings of the 2016 Annual Meeting of the International Genetic Epidemiology; 2016 October 24-26; Toronto, Canada. Genet.Epidemiol. 2016 ;40(7):656-7.

Genome-Wide Association Studies (GWAS) for complex diseases have primarily focused on the univariate analysis of each trait characterizing the disease. For example, GWAS on risk factors for coronary artery disease analyze genetic associations of plasma lipids such as total cholesterol, low-density lipoprotein, high-density lipoprotein, and triglycerides separately. However, traits underlying a disease are often correlated and a joint analysis may yield improved statistical power for association over multiple univariate analyses.We propose a unified association test of a single genetic variant with multiple traits that utilizes univariate GWAS summary statistics. This novel test does not require individual-level data, and uses only publicly available summary statistics from existing GWAS to test genetic associations of categorical and/or continuous traits. One can also use this test to analyze a single trait over multiple studies with overlapping samples. The software for the proposed test reports an approximate asymptotic P value for association and is computationally efficient for implementation at a GWAS level. Our simulation experiments show that our method can maintain proper type-I error at low error levels. It has appreciable statistical power across a wide array of association scenarios (which is unknown apriori for real data), while existing methods have widely varying power curves. When applied to plasma lipids summary data of Teslovich et al (Nature, 2010), our test detected significant genetic variants beyond the ones identified by existing tests. In summary, the proposed method can potentially provide novel insights into the genetic underpinnings of a disease.

DOI: http://dx.doi.org/10.1002/gepi.22001.

Roush GC, Amante B, Singh T, Ayele H, Araoye M, Yang D, Kostis WJ, Elliott WJ, Kostis JB, Berlin JA. Quality of meta-analyses for randomized trials in the field of hypertension: a systematic review. J.Hypertens. 2016 Dec;34(12):2305-17. PMID: 27755384.

Doubling on average every 6 years, hypertension-related meta-analyses are now published twice weekly and are often considered the highest level of evidence for clinical practice. However, some hypertension specialists and guideline authors view meta-analyses with skepticism. This article evaluates the quality of hypertension-related meta-analyses of clinical trials.
A systematic search was conducted for meta-analyses of clinical trials recently published over 3.3 years. Specific criteria reproducibly assessed 26 features in the four domains of meta-analysis quality, domains justified by fundamental analytics and extensive research: analyzing trial quality, analyzing heterogeneity, analyzing publication bias, and providing transparency.
A total of 143 meta-analyses were identified. A total of 44% had 8+ deficient features with no relation to journal impact factor: odds ratio relating 8+ deficient features to the upper third versus lower third of impact factor?=?1.3 (95% confidence limit 0.6-2.9). A total of 56% had all four domains deficient. Quality did not improve over time. Thirty articles (21%) reported statistically significant results (P?<?0.05) from inappropriate DerSimonian-Laird models, whereas unreported, appropriate, Knapp-Hartung models gave statistical nonsignificance; 88% of these 30 articles reported the incorrect results in their abstracts. A total of 60% of all meta-analyses failed to conduct analyses in subgroups of quality when indicated, 63% failed to report Tau and Tau, 57% omitted testing for publication bias, none conducted a cumulative analysis for publication bias, and 71-77% omitted mentioning in their abstracts problems of trial quality, heterogeneity, and publication bias.
Although widespread, deficiencies in hypertension-related meta-analyses are readily corrected and do not represent flaws inherent in the meta-analytic method.

DOI: https://doi.org/10.1097/HJH.0000000000001094.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/27755384.

Shim S, Kim J, Jung W, Shin IS, Bae JM. Meta-analysis for genome-wide association studies using case-control design: application and practice. Epidemiol.Health. 2016 Dec 18;38:e2016058. PMID: 28092928.

This review aimed to arrange the process of a systematic review of genome-wide association studies in order to practice and apply a genome-wide meta-analysis (GWMA). The process has a series of five steps: searching and selection, extraction of related information, evaluation of validity, meta-analysis by type of genetic model, and evaluation of heterogeneity. In contrast to intervention meta-analyses, GWMA has to evaluate the Hardy-Weinberg equilibrium (HWE) in the third step and conduct meta-analyses by five potential genetic models, including dominant, recessive, homozygote contrast, heterozygote contrast, and allelic contrast in the fourth step. The 'genhwcci' and 'metan' commands of STATA software evaluate the HWE and calculate a summary effect size, respectively. A meta-regression using the 'metareg' command of STATA should be conducted to evaluate related factors of heterogeneities.

FREE FULL TEXT: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5309730/pdf/epih-38-e2016058.pdf
DOI: https://doi.org/10.4178/epih.e2016058.
PubMed: https://www.ncbi.nlm.nih.gov/pubmed/?term=28092928.
PubMed Central: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5309730/.