Skip to main content
Figure 3 | BMC Genomics

Figure 3

From: A statistical framework for consolidating "sibling" probe sets for Affymetrix GeneChip data

Figure 3

Examples of sibling probe sets need to be consolidated (top row) and otherwise (middle and bottom rows). Examples of sibling probe sets need to be consolidated (top row) and otherwise (middle and bottom rows). In the top row, the differential expression of genes "Rho"and "Arr3" between wild type and treatment (Nrl knockout) does not change over sibling probe sets (raw P-values for interaction effect are 0.68, 0.89 respectively). In the middle and bottom rows, the differential expression between wild type and treatment for sibling probe sets have either different magnitude or reversed trend. The former is translated into a need for consolidation but not the later.

Back to article page