Skip to main content

Table 1 Estimated type I errors and proportions of censored outcomes for different sample sizes and censoring schemes (Unit: %)

From: A highly adaptive microbiome-based association test for survival traits

Method Small samples (n = 50) Large samples (n = 100)
Unif(0,10) Unif(0,5) Unif(0,10) Unif(0,5)
MiRKAT-S(KU) 4.97 4.99 4.94 4.99
MiRKAT-S(K0.5) 4.99 4.98 5.10 5.03
MiRKAT-S(KW) 5.00 4.92 5.09 5.06
MiRKAT-S(KBC) 4.98 4.93 5.00 4.99
OMiRKAT-S 4.97 4.91 5.06 5.12
MiSALN(1/4) 4.99 4.96 4.91 5.13
MiSALN(1/3) 5.04 4.95 4.93 5.13
MiSALN(1/2) 4.98 4.92 4.88 5.06
MiSALN(1) 4.96 4.93 4.90 5.03
OMiSALN 5.01 4.93 4.92 5.07
OMiSA 4.94 4.94 4.98 5.15
Prop. of Censoring 25.78 40.42 25.88 40.48
  1. KU, K0.5, KW, and KBC, indicates the use of unweighted UniFrac, generalized UniFrac with ϴ = 0.5, weighted UniFrac, and the Bray-Curtis dissimilarity kernels, respectively, for MiRKAT-S [24]
\