Table 3 The contingency tables between the live top scoring attributes given in Fig. 2, along with the z-score S for the pair-wise mutual information I(A _i , A _j ), estimated according to Eq. (8) with respect to 500 randomized realizations. High values of S indicate that both attributes are not independent, i.e. that the probability of observing such a value of the mutual information I(A _i , A _j ) for statistically independent attributes A _i and A _j is low. Shown are the nodes: 'GO:0005830' (component: cytosolic ribosome). 'GO:0003735' (function: structural constituent of ribosome). 'GO:0005840' (component: ribosome). 'GO:0006412' (process: protein biosynthesis), and 'GO:0019538' (process: protein metabolism). Note that the contingency tables, as well as the z-score, was estimated for the full set of 6312 genes. Reducing the analysis to those 3000 genes used in the creation of Fig. 2 increases the redundancy even more.

GO:0005830

1

140

0

137

3

140

0

137

3

137

3

0

0

6172

61

6111

89

6083

421

5751

709

5463

-

S ≈ 702

S ≈ 687

S ≈ 430

S ≈ 394

GO:0003735

1

137

61

198

0

196

2

198

0

198

0

0

3

6111

0

6114

33

6081

360

5754

648

5466

-

-

S ≈ 1037

S ≈ 704

S ≈ 585

GO:0005840

1

140

89

196

33

229

0

219

10

220

9

0

0

6083

2

6081

0

6083

339

5744

626

5457

-

-

-

S ≈ 771

S ≈ 649

GO:0006412

1

137

421

198

360

219

339

558

0

558

0

0

3

5751

0

5754

10

5744

0

5754

288

5466

-

-

-

-

S ≈ 1663

GO:0019538

1

137

709

198

648

220

626

558

288

816

0

0

3

5463

0

5466

0

5457

0

5466

0

5466

-

-

-

-

-