|
|
|
|
|
Validation and refinement of gene-regulatory pathways on a network of physical interactions
|
A major challenge of systems biology is to elucidate molecular networks that control cellular function. Towards this goal, we have assembled a catalog of 38 physical network models constructed from a scaffold of protein-protein and protein-DNA interactions using a causal probabilistic approach to explain changes in gene expression observed across 273 publicly-available gene knockout experiments in yeast. The models encode a variety of previously unknown regulatory mechanisms, such as four distinct pathways explaining extensive expression changes downstream of Sok2 and the repression of Gcn4 by Gln3 through direct transcriptional control. In addition, to refine these models we present a novel automated procedure for selecting new knockout experiments that reveal the directionality of each interaction and its regulatory role as an activator or inhibitor. The observed knock-out effects from implied experiments are consistent with our model predictions and suggest that automated model building and experimental design are effective strategies for characterizing information flow across biological networks.
|
|
 |
| Information Flow Induced in swi4 deletion
|
 |
|
 |
|
A model of observed expression responses demonstrating regulatory pathways downstream of Swi4 and Sok2 after a swi4 deletion. More information, including legends, is available in the Data Download section. Diagram visualization was performed using the Cytoscape network modeling environment (http://www.cytoscape.org/).
|
|
|
|
|
|
|
|
|
|
|
|
All logos and trademarks in this site are property and (c) 2005 by the Laboratory for Integrative Network Biology, UC San Diego Department of Bioengineering.
