dN/dS of kernels is significantly lower than plug-ins, I/Os and
batteries, see Table 2 for P value. Also, we observe that the regulatory
gene group of kernels and plug-ins has a lower dN/dS value than
group of I/Os and batteries (relative difference=−0.055, P
value=0.0157 for Wilcoxon test). From the probability distribution
of dN/dS in Fig. 4, we can see the distribution of kernels is narrow
width, and the peak probability appears at a low dN/dS. If only TF
genes are considered, kernels (dN/dS=0.045) still evolve more
slowly than other components. We also see slight increase of dN/dS
from plug-ins to I/Os and to batteries (dN/dS=0.138). Interestingly,
we found that the number of organisms for which an ortholog was
detected varies from genes to genes. For example, A. miniata is the
least close organism to S. purpuratus compared to other sea urchins, S.
purpuratus and A. miniata are in the same phylum but different orders.
We found four orthologous genes between A. miniata and S.
purpuratus. Three of them are kernel genes. The orthologs of kernel
genes are more likely detected than other genes in far related
organism, which is a support of slower evolution of kernels. Our
results show that if two organisms are in the same phylum, their
kernel modules that determine the phylum-level body plan are
conserved. If two organisms are in the same class or order, their plugins
and I/O modules are conserved since they determine the class and
order level body plan.