 
R

* ∑(d_{+}d_{})

∑c

* ∑(d_{+}d_{}c)

$\frac{{\displaystyle \sum ({d}_{+}{d}_{})}}{R}$

$\frac{{\displaystyle \sum {c}_{i}}}{R}$

$\frac{{\displaystyle \sum ({d}_{+}{d}_{}c)}}{R}$

$\sum}_{\mathcal{I}}{d}_{i$

${\sum}_{\mathcal{O}}d$

$\left({\displaystyle {\sum}_{\mathcal{I}}d}\right)\left({\displaystyle {\sum}_{\mathcal{O}}d}\right)$


r
Prelog

0.459

0.656

0.509

0.666

0.6

0.0178

0.62

0.135

0.604

0.408

r
Postlog

0.827

0.875

0.764

0.870

0.860

0.343

0.856

0.426

0.561

0.623

ρ

0.841

0.876

0.943

0.876

0.855

0.059

0.845

0.478

0.496

0.603

Range

Min

8

15

5

2

1.67

0.09

0.06

2

2

6


Max

174

89132

35.08

5781.73

665.38

0.9

47.78

25

47

414

 A good factor for an estimating function must have a high correlation to that is being estimated. We further required that the factor must grow consistently with the number of ExPas. The rows labelled 'Pre' and 'Post' Log show the Pearson's correlations, r, between the number of ExPas and the corresponding factors. These factors were created using the following basic networks measurements: R = R_{
eff
}the number of active reactions given the environmental conditions, d_{±} = d_{±}(r_{
i
}) the incoming/outgoing connectivity of reaction r_{
i
}, c = c(r_{
i
}) the clustering coefficient of the i^{th}reaction, $\mathcal{I}$ the set of input reactions, and $\mathcal{O}$ the set of output reactions. Both Pearson's and Spearman's Rank correlation coefficients, r and ρ were used a guide to identify reliable contributing factors. Given this information, the final chosen factors are emphasized by an asterisk (*).