Skip to main content

Table 6 Accuracy measures: Illumina's default method versus rmx estimator

From: Preprocessing of gene expression data by optimally robust estimators

Estimator rmx Illumina Mean median 5% trim 5% winsorize
25% SD 0.114 0.116 0.125 0.121 0.117 0.122
median SD 0.133 0.135 0.140 0.142 0.133 0.137
75% SD 0.150 0.153 0.157 0.160 0.150 0.153
99% SD 0.308 0.309 0.314 0.310 0.310 0.313
null log-fc IQR 0.186 0.187 0.208 0.208 0.188 0.192
null log-fc 99% 0.362 0.366 0.412 0.408 0.377 0.388
null log-fc 99.9% 0.495 0.504 0.584 0.555 0.525 0.546
  1. Accuracy measures of [11] and [12] (smaller is better) for the rmx estimator for s [0, 0.05], Illumina's default method as well as mean and SD (mean), median and MAD (median), 5% trimmed mean and SD (5% trim) and 5% winsorized mean and SD (5% winsorize).