A novel application of quantile regression for identification of biomarkers exemplified by equine cartilage microarray data

Table 1 Transforming quantiles of log₂(R/G) to Z score.

Quantile of log2(R/G)	Z-score
observed log₂(R/G) ≥ 99^th estimated quantile	2.57
99^th estimated quantile > observed log₂(R/G) ≥ 95^th estimated quantile	1.96
95^th estimated quantile > observed log₂(R/G) ≥ 90^th estimated quantile	1.44
90^th estimated quantile > observed log₂(R/G) ≥ 80^th estimated quantile	1.04
80^th estimated quantile > observed log₂(R/G) ≥ 50^th estimated quantile	0.39
50^th estimated quantile > observed log₂(R/G) ≥ 20^th estimated quantile	-0.39
20^th estimated quantile > observed log₂(R/G) ≥ 10^th estimated quantile	-1.04
10^th estimated quantile > observed log₂(R/G) ≥ 5^th estimated quantile	-1.44
5^th estimated quantile > observed log₂(R/G) ≥ 1^stestimated quantile	-1.96
observed log₂(R/G) < 1^st estimated quantile	-2.57

ISSN: 1471-2105