Calibration of credibility of agnostic genome-wide associations
Genome-wide testing platforms are increasingly used to promote "agnostic" approaches to the discovery of gene variants associated with the risk of many common diseases and quantitative traits. The early track record of genome-wide association (GWA) studies suggests that some proposed associations are replicated quite consistently with large-scale subsequent evidence from multiple studies, others have a more inconsistent replication record, some have failed to be replicated by independent investigators and many more early proposed associations await further replication. An important question is how to calibrate the credibility of these postulated associations. A simple Bayesian method is applied here to achieve such calibration. The variability of the estimated credibility is examined under different assumptions. Empirical examples are drawn from existing GWA studies. It is demonstrated that the credibility of different proposed associations can cover a very wide range. The credibility of specific associations usually remains relatively robust when different plausible assumptions are made (within a reasonable range) for the prior odds of an association being true, or the magnitude of the anticipated effect size for genetic associations. Heterogeneity and bias assumptions can have a more major impact on the credibility estimates and thus they need very careful consideration in each case. Credibility calibration may be used in conjunction with qualitative criteria for the appraisal of the cumulative evidence that take into consideration the amount, consistency, and protection from bias in the data.
Available from: Lars Bertram
- "Empirical analyses of the AlzGene database, which uses the epidemiological approach of the Venice criteria (evidence , consistency and bias), show that when using a p value threshold of 0.05 as a threshold for ''significance'' upon meta-analysis of candidate genes in at least four independent datasets, still more than half of the findings (12/21) will turn out to be false. The finding that a p value threshold of 0.05 yields a low PPV when the prior probability of association is low corroborates findings of others [22, 23], and provides one explanation as to why many candidate gene studies using this p value threshold were not replicated. This finding also extends to non-genetic studies of epidemiological risk factors. "
[Show abstract] [Hide abstract]
ABSTRACT: Distinguishing true from false positive findings is a major challenge in human genetic epidemiology. Several strategies have been devised to facilitate this, including the positive predictive value (PPV) and a set of epidemiological criteria, known as the "Venice" criteria. The PPV measures the probability of a true association, given a statistically significant finding, while the Venice criteria grade the credibility based on the amount of evidence, consistency of replication and protection from bias. A vast majority of journals use significance thresholds to identify the true positive findings. We studied the effect of p value thresholds on the PPV and used the PPV and Venice criteria to define usable thresholds of statistical significance. Theoretical and empirical analyses of data published on AlzGene show that at a nominal p value threshold of 0.05 most "positive" findings will turn out to be false if the prior probability of association is below 0.10 even if the statistical power of the study is higher than 0.80. However, in underpowered studies (0.25) with a low prior probability of 1 × 10(-3), a p value of 1 × 10(-5) yields a high PPV (>96 %). Here we have shown that the p value threshold of 1 × 10(-5) gives a very strong evidence of association in almost all studies. However, in the case of a very high prior probability of association (0.50) a p value threshold of 0.05 may be sufficient, while for studies with very low prior probability of association (1 × 10(-4); genome-wide association studies for instance) 1 × 10(-7) may serve as a useful threshold to declare significance.
Available from: Jean-Charles Lambert
- "Power would be only 20% for a minor allele frequency of 2% and odds ratio of 1.15, but it would be 78% for the same minor allele frequency of 2% and odds ratio of 1.3. Furthermore , for associations with uncorrected p < 0.05, we also expressed the strength of the putative associations by calculating the respective Bayes factor, assuming that average genetic effects for PD susceptibility with common variants may reflect odds ratios of 1.30 and using a lump-and-smear prior, as described by Ioannidis (Ioannidis, 2008). Meta-analyses were performed using Review Manager 4.2.7 and STATA 10.0 (Stata Corp., College Station, TX, USA). "
[Show abstract] [Hide abstract]
ABSTRACT: High-profile studies have provided conflicting results regarding the involvement of the Omi/HtrA2 gene in Parkinson's disease (PD) susceptibility. Therefore, we performed a large-scale analysis of the association of common Omi/HtrA2 variants in the Genetic Epidemiology of Parkinson's disease (GEO-PD) consortium. GEO-PD sites provided clinical and genetic data including affection status, gender, ethnicity, age at study, age at examination (all subjects); age at onset and family history of PD (patients). Genotyping was performed for the five most informative SNPs spanning the Omi/HtrA2 gene in approximately 2-3 kb intervals (rs10779958, rs2231250, rs72470544, rs1183739, rs2241028). Fixed as well as random effect models were used to provide summary risk estimates of Omi/HtrA2 variants. The 20 GEO-PD sites provided data for 6378 cases and 8880 controls. No overall significant associations for the five Omi/HtrA2 SNPs and PD were observed using either fixed effect or random effect models. The summary odds ratios ranged between 0.98 and 1.08 and the estimates of between-study heterogeneity were not large (non-significant Q statistics for all 5 SNPs; I(2) estimates 0-28%). Trends for association were seen for participants of Scandinavian descent for rs2241028 (OR 1.41, p=0.04) and for rs1183739 for age at examination (cut-off 65 years; OR 1.17, p=0.02), but these would not be significant after adjusting for multiple comparisons and their Bayes factors were only modest. This largest association study performed to define the role of any gene in the pathogenesis of Parkinson's disease revealed no overall strong association of Omi/HtrA2 variants with PD in populations worldwide.
Available from: aje.oxfordjournals.org
Data provided are for informational purposes only. Although carefully collected, accuracy cannot be guaranteed. The impact factor represents a rough estimation of the journal's impact factor and does not reflect the actual current impact factor. Publisher conditions are provided by RoMEO. Differing provisions from the publisher's actual policy or licence agreement may be applicable.