Skip to main content
. 2016 Sep 6;34(32):3931–3938. doi: 10.1200/JCO.2016.68.1031

Fig A1.

Fig A1.

Box plots of the estimated handling effects for 192 arrays in the nonuniformly handled data set. Array allocation to the training set is indicated by the shading: white for the training set and gray for not used. Below the x-axis are labels for the array slide (indexed as 1-24), handling batch (indexed as 1-5), handling technician (indexed as 1 and 2), sample group assignment in the empirical data (E for endometrial; V for ovarian), and sample group assignment for the simulated designs: Complete Confounding 1 (CC1) and Complete Counfounding 2 (CC2).