Limits...
Power and sample size estimation for epigenome-wide association scans to detect differential DNA methylation.

Tsai PC, Bell JT - Int J Epidemiol (2015)

Bottom Line: We performed simulations to estimate power under the case-control and discordant MZ twin EWAS study designs, under a range of epigenetic risk effect sizes and conditions.Our analyses highlighted several factors that significantly influenced EWAS power, including sample size, epigenetic risk effect size, the variance of DNA methylation at the locus of interest and the correlation in DNA methylation patterns within the twin sample.Our results can help guide EWAS experimental design and interpretation for future epigenetic studies.

View Article: PubMed Central - PubMed

Affiliation: Department of Twin Research and Genetic Epidemiology, King's College London, London, UK.

No MeSH data available.


Power of discordant twin EWAS. Estimates are shown for the twin (solid lines) and case-control (dashed lines) designs for a range of sample sizes and mean differences at a significance level of 0.05 (A, upper panel) and 1 × 10−6 (B, lower panel). Each line represents the power curve under different sample sizes from 10 to 100 pairs of twins, or pairs of cases and controls.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC4588864&req=5

dyv041-F4: Power of discordant twin EWAS. Estimates are shown for the twin (solid lines) and case-control (dashed lines) designs for a range of sample sizes and mean differences at a significance level of 0.05 (A, upper panel) and 1 × 10−6 (B, lower panel). Each line represents the power curve under different sample sizes from 10 to 100 pairs of twins, or pairs of cases and controls.

Mentions: We next estimated EWAS power under the disease-discordant MZ twin design. Simulations were performed with mean difference effects from 1% to 60% and with sample sizes of 10, 15, 20, 25, 30 and 50 twin pairs (Table 1, Figure 4). For example, we observed that a sample of 25 twin pairs has over 80% power to detect a mean difference of 8% in methylation at nominal significance (P = 0.05), and 25% at genome-wide significance (P = 1 × 10−6). As expected, power estimates in twins outperformed the case-control design (Table 1, Figure 4). For example, a sample of 25 twin pairs has over 80% power to detect a mean difference of 8% in methylation at nominal significance (P = 0.05), whereas 25 pairs of cases and controls have only 45% power to detect this effect (Figure 4A). At genome-wide significance, at least 50 pairs of subjects were required to identify effect sizes of 16% mean difference with over 80% power in both designs (Figure 4B). However, our simulations were not designed for a formal comparison between case-control and twin power, because our results assume that twins and case-control samples are equally well matched for factors that can influence differential methylation, including age, sex and cohort effects, and unrelated samples are typically more heterogeneous than MZ twins.Figure 4.


Power and sample size estimation for epigenome-wide association scans to detect differential DNA methylation.

Tsai PC, Bell JT - Int J Epidemiol (2015)

Power of discordant twin EWAS. Estimates are shown for the twin (solid lines) and case-control (dashed lines) designs for a range of sample sizes and mean differences at a significance level of 0.05 (A, upper panel) and 1 × 10−6 (B, lower panel). Each line represents the power curve under different sample sizes from 10 to 100 pairs of twins, or pairs of cases and controls.
© Copyright Policy - creative-commons
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC4588864&req=5

dyv041-F4: Power of discordant twin EWAS. Estimates are shown for the twin (solid lines) and case-control (dashed lines) designs for a range of sample sizes and mean differences at a significance level of 0.05 (A, upper panel) and 1 × 10−6 (B, lower panel). Each line represents the power curve under different sample sizes from 10 to 100 pairs of twins, or pairs of cases and controls.
Mentions: We next estimated EWAS power under the disease-discordant MZ twin design. Simulations were performed with mean difference effects from 1% to 60% and with sample sizes of 10, 15, 20, 25, 30 and 50 twin pairs (Table 1, Figure 4). For example, we observed that a sample of 25 twin pairs has over 80% power to detect a mean difference of 8% in methylation at nominal significance (P = 0.05), and 25% at genome-wide significance (P = 1 × 10−6). As expected, power estimates in twins outperformed the case-control design (Table 1, Figure 4). For example, a sample of 25 twin pairs has over 80% power to detect a mean difference of 8% in methylation at nominal significance (P = 0.05), whereas 25 pairs of cases and controls have only 45% power to detect this effect (Figure 4A). At genome-wide significance, at least 50 pairs of subjects were required to identify effect sizes of 16% mean difference with over 80% power in both designs (Figure 4B). However, our simulations were not designed for a formal comparison between case-control and twin power, because our results assume that twins and case-control samples are equally well matched for factors that can influence differential methylation, including age, sex and cohort effects, and unrelated samples are typically more heterogeneous than MZ twins.Figure 4.

Bottom Line: We performed simulations to estimate power under the case-control and discordant MZ twin EWAS study designs, under a range of epigenetic risk effect sizes and conditions.Our analyses highlighted several factors that significantly influenced EWAS power, including sample size, epigenetic risk effect size, the variance of DNA methylation at the locus of interest and the correlation in DNA methylation patterns within the twin sample.Our results can help guide EWAS experimental design and interpretation for future epigenetic studies.

View Article: PubMed Central - PubMed

Affiliation: Department of Twin Research and Genetic Epidemiology, King's College London, London, UK.

No MeSH data available.