Table 2

Inflation before and after principal component correction

Model

<a onClick="popup('http://www.biomedcentral.com/1471-2156/11/80/mathml/M1','MathML',630,470);return false;" target="_blank" href="http://www.biomedcentral.com/1471-2156/11/80/mathml/M1">View MathML</a>without PC adjustment

<a onClick="popup('http://www.biomedcentral.com/1471-2156/11/80/mathml/M1','MathML',630,470);return false;" target="_blank" href="http://www.biomedcentral.com/1471-2156/11/80/mathml/M1">View MathML</a>with PC adjustment


Single-SNP tests*

1.070

1.077

Main-effects

1.101

1.059

Main+adj

1.092

1.052

Haplotype

1.064

1.037


Comparison of inflation in median test statistics (<a onClick="popup('http://www.biomedcentral.com/1471-2156/11/80/mathml/M1','MathML',630,470);return false;" target="_blank" href="http://www.biomedcentral.com/1471-2156/11/80/mathml/M1">View MathML</a>) for each model with and without principal components. *Single-SNP tests without PC adjustment were Cochran-Armitage trend statistics which are not able to incorporate PC adjustment. The single-SNP statistics with PC adjustment are the 1 d.f. logistic regression.

Wason and Dudbridge BMC Genetics 2010 11:80   doi:10.1186/1471-2156-11-80

Open Data