Li Chen

4/3/2009

CSc 8910 Analysis of Biological Network, Spring 2009

Dr. Yi Pan

Introduction Results Conclusions

Transcriptional Regulatory Network A complex network of interactions among transcription

factors and promoter regions of genes and operons. Goal of Identifying Motifs in Transcriptional

Regulatory Network To simplify networks’ architecture and better

understand the system-level function of such networks.

Previous achievement The motifs could be identified in the network. But they

are small, overrepresented, topologically distinct regulatory interaction patterns.

First organizational level: motifs Each network being characterized by its

own set of distinct motifs. In the E.coli transcriptional regulatory

network, majority of motifs are feed-forward motifs and bi-fan motifs.

Feed-forward and bi-fan motifs can be classified by the functionality of their links, namely, activating or inhibitory .

• (1) coherent type feed-forward motif (FF)• (2) incoherent type feed-forward motif (FF)• (3) coherent bi-fan motif (BF)• (4) incoherent bi-fan motif (BF) Graphical representation of the network.

Blue diamonds: transcription factors (TF)

Red circles: regulated operons

Links: blue -- activatorgreen -- repressorbrown -- activator or repressor effect

Detailed statistics of the nodes (upper table) and the two statistically significant motifs (bottom table) found in the network.

Second organizational level: homologous motif clusters Feed-forward motifs that share at least one link and/or node

with another feed-forward motif. Forty-one of the 42 individual feed-forward motif clusters

Six motif clusters, three have

one highly shared link, while a a shared node plays a critical role in establishing the other three motif clusters.

Bi-fan motifs that share at least one link and/or node with another bi-fan motif.

208 of the 209 bi-fan motifs

join together into just two bi-

fan motif clusters.

Most of links are shared by at

least two adjacent motifs, and

also among multiple motifs.

Third organizational level: motif super-cluster

Merge all feed-forward and bi-fan

homologous motif clusters

Form a single large connected component (motif super-cluster)

Vast majority of feed-forward motif clusters share the same links with the bi-fan motif clusters

The relationship of organizational levels to the global network topology The connected giant component of the complete E.coli

transcriptional regulatory

Cumulative connectivity distribution P(k) The solid

black line has an

exponent γ=-1.5,

provides the best fit

for the original

network (black

circles)

The clustering distribution C(k) The solid black line has slope ζ = -1, and the is the best fit for all networks. The clustering

coefficient of a node is a measure of its near-neighbors connectivity, and thus for the BF motifs this value is zero.

Demonstrate the heterologous motif super-cluster represents the backbone of the connected giant component• Removing all 250 links of super-cluster from the network.

• Removal of 250 randomly chosen links. Network break

into 16 small sub-graphs, a

connected giant component was

retained.

The connectivity distribution P(k) of the remaining networks The solid line has slope

γ = -2, and is the best fit for the random link removal.

After random link removal, P(k) is relatively unaltered, being reminiscent to that

observed for the original network.

The clustering distribution C(k) The solid line has

slope ζ = -2. After super-

cluster links removal, C(k)

and k was completely

absent.

For the E. coli transcriptional regulatory network, Individual motifs , homologous motif clusters and super-cluster are key determinants of the network’s global topological organization.

Individual motifs, homologous motif clusters and super-cluster may represent distinct organizational hierarchies of transcriptional regulation.

It is likely that the aggregation of motifs into motif clusters and super-clusters is a general property of all cellular networks.