Table 2 Array outliers.

Sample	pos_vs_neg_auc	all_probeset_mean	all_probeset_rle_mean	Sample	pos_vs_neg_auc	all_probeset_mean	all_probeset_rle_mean
11SE560	0.85422	5.541424	0.241511	11SE576	0.864426	5.589662	0.15491
11SE561	0.878029	5.57988	0.171301	11SE577	0.871134	5.576477	0.129484
11SE562	0.865219	5.507742	0.248986	11SE578	0.88545	5.576419	0.128479
11SE563	0.864947	5.589251	0.175722	11SE579	0.866908	5.576636	0.127985
11SE564	0.864352	5.5837	0.173877	11SE580	0.857355	5.586262	0.153071
11SE565	0.877761	5.56282	0.161048	11SE581	0.867117	5.574453	0.131372
11SE566	0.871477	5.586922	0.175144	11SE582	0.88222	5.572265	0.130127
11SE567	0.877053	5.580364	0.167735	11SE583	0.874066	5.581122	0.141788
11SE568	0.860977	5.579972	0.140385	11SE584	0.880905	5.571299	0.122088
11SE569	0.880092	5.582358	0.217311	11SE585	0.846603	5.579253	0.170314
11SE570	0.882	5.574604	0.148582	11SE586	0.876601	5.576181	0.131724
11SE571	0.888591	5.56232	0.194358	11SE587	0.872058	5.566449	0.226278
11SE572	0.876405	5.570266	0.133905	11SE588	0.904137	5.559667	0.184642
11SE573	0.879322	5.56958	0.12135	11SE589	0.873042	5.570007	0.150338
11SE574	0.875899	5.572968	0.116654	11SE590	0.842198	5.593665	0.170785
11SE575	0.863548	5.58026	0.129136	11SE591	0.879472	5.567289	0.125395

Array outliers. Pos_vs_neg_auc is a measure of the detection of the positive controls versus false detection of the negative controls. It is a robust measure of the global quality of the data. Typical values range 0.8–0.9, being 1.0 the perfect value. In contrast, 0.5 indicates a lack of difference between positive and negative controls. Values <0.8 indicate that the sample may be an outlier. All Probe set Mean is a measure of the average signal of all probe sets included in the analysis. It allows the identification of arrays too bright or too dark. All Probe Set RLE Mean: the signal from each probe set is compared with the average signal of that probe set in the study. This parameter is the average of those differences for all the probe sets. Generally, high values indicate that the signals in an array are very different from the rest of the signals of the study. In this kind of array, values usually range 0.27–0.61 for studies that include samples from various tissues, and 0.1–0.23 for technical replicates. Abbreviations: RLE, relative log expression.

Search