Table 3.
Results of the leave-one-out sensitivity analysis. For each contrast, an individual gene is listed if its removal produced a change in the ranking of the six methodologies. The third column shows how the ranking of the six methodologies changed upon removal of the gene. Here, M = Mas5, G = gcRMA, R = RMA, V = VSN, D = dChip, D- = dChip.mm. Bold font highlights changes. Note that all changes in rankings, with two exceptions, were transpositions of two adjacent methods or a shuffle of three adjacent methods.
CONTRAST | GENE | RANKING |
Age | 36 | (M,D-,G,R,V,D)→(G,M,D-,R,V,D) |
Genotype | 3 | (G,M,D-,R,V,D)→(D-,G,M,R,V,D) |
Genotype | 12 | (G,M,D-,R,V,D)→(M,G,D-,V,R,D) |
Genotype | 26 | (G,M,D-,R,V,D)→(G,M,R,D-,V,D) |
Genotype | 27 | (G,M,D-,R,V,D)→(G,D-,M,R,V,D) |
Interaction | 4,19 | (M,D-,G,R,D,V)→(M,G,D-,R,D,V) |
Interaction | 6,11,24 27,34,37,43,47 | (M,D-,G,R,D,V)→(D-,M,G,R,D,V) |
Interaction | 8 | (M,D-,G,R,D,V)→(M,D-,R,G,D,V) |
Interaction | 26 | (M,D-,G,R,D,V)→(G,M,D-,R,D,V) |
Interaction | 32 | (M,D-,G,R,D,V)→(M,D-,G,R,V,D) |
Interaction | 12 | (M,D-,G,R,D,V)→(G,D-,R,M,V,D) |
YWT-OWT | 3 | (M,G,D,R,V,D-)→(M,G,D,R,D-,V) |
YWT-OWT | 37 | (M,G,D,R,V,D-)→(G,M,D,R,V,D-) |
YWT-YMCAT | 3 | (M,G,D,R,V,D-)→(M,D,G,R,V,D-) |
OWT-OMCAT | 2,4,8,15,16,20,27,31,34,35,43,47 | (D-,G,V,M,R,D)→(D-,G,V,R,M,D) |
OWT-OMCAT | 10,12,17,18,21,36,39,40,45 | (D-,G,V,M,R,D)→(D-,G,M,V,R,D) |
OWT-OMCAT | 23 | (D-,G,V,M,R,D)→(D-,V,G,R,M,D) |
OWT-OMCAT | 5,6,13,32,33 | (D-,G,V,M,R,D)→(D-,G,R,M,V,D) |
OWT-OMCAT | 3 | (D-,G,V,M,R,D)→(D-,M,G,V,R,D) |
OWT-OMCAT | 26 | (D-,G,V,M,R,D)→(G,R,M,V,D-,D) |