SEARCH

SEARCH BY CITATION

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

FilenameFormatSizeDescription
gepi21665-sup-0001-FigureS1.doc164KFigure S1 Null distribution computed numerically from 106 replicates of a sample of 2000 cases and 2000 controls. Subplots (A) and (B) are computed under conditions 1 and 2, respectively (see text). The red line represents a chi square distribution with 4 degrees of freedom.
gepi21665-sup-0002-FigureS2.doc100KFigure S2 Probability-probability density plot for PZ (x axis) and LRIT method (y axis). The color represents the number of tests on a logarithmic scale in each point determined by the P-values obtained by both methods for each data set for the 48 basic epistatic models. Although LRIT is marginally more powerful, it is clear that the performance of both methods is very similar.
gepi21665-sup-0003-FigureS3.doc226KFigure S3 Probability-probability density plot for the case-only χ2 (x axis) and the case-control 2χ2 method (y axis). The color represents the number of tests on a logarithmic scale in each point determined by the P-values obtained by both methods for each data set for the 48 basic epistatic models. Different prevalences (0.01, 0.05, 0.1, 0.15, 0.2) are colored with different gradient colors. Higher prevalences present higher slope. The simulations are done with 500 cases and 500, 1000, and 1500 controls for subplots (A), (B), and (C), respectively.

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.