Fig. 1From: Longitudinal linear combination test for gene set analysisCalculation of the power of LLCT using simulated data generated with different within-geneest correlation (a), sample size (b), geneset size (c), number of repeated measurements (d) and within-subject correlation (e). Type I error is set at 5%. For each plot, the simulation variables except the one mentioned on the title varies but remains comparable among the curvesBack to article page