Disclaimer: Supplementary materials have been peer-reviewed but not copyedited.


Fig. S1. Power comparison of two-marker versus single-marker tests in the first simulation. a) Scatter plot of power of the two-marker test versus the single-marker test (with 40×40×40 correlation combinations). b) Histogram of power difference between the two-marker test and the single-marker test.

Fig. S2. Power comparison for different causal allele RRs, sample sizes (ssize) and Type I error rates. The y axis is the proportion of cases where two-marker tests have greater power than single-marker test for a fixed ρ12. Each plot represents a set of RRs and a given sample size. The green, blue and red lines correspond to α= 0.05, 0.001 and 5×10-8, respectively.

Fig. S3. Surface plot of the distance between the two turning points (shown in Fig 1c) under a range of standardized effect sizes and type I error rates used in the second simulation.

Fig. S4. The observed space of correlations (ρ12 is fixed around 0.5) from the simulated SNPs when the number of possible haplotypes was restricted to three or four.

Please note: Wiley Blackwell is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.