Skip to main content

Table 2 Empirical type I error rate of unadjusted and covariate-adjusted KRV at nominal level \(\alpha\) under Type I Error Scenario 1

From: Kernel-based genetic association analysis for microbiome phenotypes identifies host genetic drivers of beta-diversity

Method

Microbiome kernel

\(\alpha\)

  

0.05

0.01

0.001

Unadjusted KRV

Bray-Curtis

0.2403

0.0936

0.0255

 

Unweighted UniFrac

0.0484

0.0094

0.0011

 

Weighted UniFrac

0.1371

0.0371

0.0057

 

Generalized UniFrac

0.1412

0.0416

0.0063

 

CLR-linear

0.0811

0.0178

0.0016

 

PhILR-linear

0.1389

0.0434

0.0076

Adjusted KRV

Bray-Curtis

0.0473

0.0114

0.0012

 

Unweighted UniFrac

0.0523

0.0115

0.0009

 

Weighted UniFrac

0.0507

0.0095

0.0012

 

Generalized UniFrac

0.0499

0.0097

0.0011

 

CLR-linear

0.0450

0.0091

0.0011

 

PhILR-linear

0.0482

0.0093

0.0015

  1. Linear kernel was used for genetic data
Back to article page

Microbiome

ISSN: 2049-2618

Contact us