Table 2.

Parameters for the power-law distribution (b, k) for (A) HGT vine widths (Fig. 1) and (B) the connectivity of the network (Fig. 2), according to the four methods used

y = a * xk; where a = exp(b) = eb
Method
b
K
R2
A
   Average ortholog similarity 11.8 —2.88 0.95
   Gene content 11.8 —2.93 0.96
   Genome conservation 11.7 —2.84 0.94
   STRING 9.9 —2.68 0.95
B
   Average ortholog similarity 6.7 —1.93 0.68
   Gene content 7.4 —2.25 0.77
   Genome conservation 5.8 —1.54 0.72
   STRING
6.0
—2.55
0.83

[i] Goodness-of-fit is expressed as the coefficient of determination (R2) defined as R2 = 1–SSE/SSM, where SSE is the sum of squared errors, and SSM is the sum of squares around the mean.