Schema for regulatory network
graphs showing the types of nodes (every graph node must be one of
these types) and the allowed connections between them. In the schema,
there are three Central Dogma nodes for gene, mRNA and protein, along
with suitable relationships. These nodes are augmented with
transcriptional regulatory elements at the single-site and
module/cassette level. Both protein:protein complexes and protein:DNA
complexes receive their own node types. There are also three nodes
types for reactions involving protein: degradation, assembly, and
state modification such as ubiqitination, neddylation, or
phosphorylation. Each of these could have been drawn as a kind of
link rather than node, except that we know they often involve three or
more protein components and it is hard to draw or compute with
three-headed arrows. We therefore transform them to a standard
bipartite graph representation as a simple way to handle the problem.
This design will enable us to produce large regulatory networks as an
“unrolled” version of this schema, in which each schema node has many
instances and each schema link (between nodes of type A and B, let us
say) corresponds to many actual links between individual nodes of type
A and B. Such a network is consistent with the schema of the Figure
above.. |