Improving Boolean Networks to Model Signaling Pathways Bree

Improving Boolean Networks to Model Signaling Pathways Bree Aldridge Diana Chai BE. 400 Term Project December 5, 2002

Outline Motivation / Project Goals Introduction to Model System Implementation: Boolean network Fuzzy network Results / Conclusions Future Work

Motivation Cellular states control behavior Quantitative signaling and state data difficult to obtain Boolean-like networks: Representative of how signaling networks process and transmit information “Simpler” than solving a huge system of ODEs Tool to explore subnetwork interactions (crosstalk) Missing data holes may be filled in with intuition

Project Goals Explore the use of Boolean-like networks to model signaling events Determine level of abstraction to which Boolean-like networks are useful Make qualitative predictions about important nodes in signaling pathways

Model System Fibronectin a 5 b 1 Insulin Grb 2 Insulin Receptor FAK/Src Sos IRS 1 P 13 K Ras Akt/PKB Raf Mek Erk DNA Synthesis Asthagiri and Lauffenburger, 2001 Anabi et al. , 2001

Transient Behavior Asthagiri and Lauffenburger, 2001

DNA Synthesis Asthagiri et. al. , 2000

Pure Boolean Model

Pure Boolean Output

Fuzzified Model Go to Simulink: Introduction to fuzzy logic Membership functions Rule based logic Show working model

Results : Time course

Results: DNA Synthesis

Take-home Results Fuzzy logic networks are capable of capturing qualitative features of signaling networks (e. g. crosstalk) Easy to build despite lack of quantitative information Good for testing hypotheses at higher level of abstraction than ODE-based models

Crosstalk Example

Conclusions Boolean Networks are NOT sufficient to capture complex behaviors of signalling networks where behavior is not ALL or NONE Fuzzy Logic Networks are best used at the qualitative prediction level Also good for exploring how subnetworks interact Especially good for when data is lacking

Future Work Explore the insulin signaling pathway Explore different levels of crosstalk Explore sensitivity by changing membership functions and weighting rules

References Annabi, Gautier, and Baron, Fed. Eur. Biochem. Soc. , 507, 247 -252 (2001) Assoian and Schwartz, Curr. Opin. Genet. Dev. 11, 48 -53 (2001) Asthagiri and Lauffenburger, Biotechnol. Prog. 17, 227 -239 (2001) Asthagiri, Reinhart, Horwitz, and Lauffenburger, J. Cell Sci. , 113, 44994510 (2000) Asthagiri et. al. , J. Biol. Chem. , 274, 27119 -27127 (1999) Eliceiri, Circ. Res. , 89, 1104 -1110 (2001) Giancotti and Ruoslahti, Science 285, 1028 -1032 (1999) Guilherme , Torres, and Czech, J. Biol. Chem. , 273, 22899 -22903 (1998) Huang and Ferrell, PNAS, 93, 10078 -10083 (1996) Huang and Ingber, Exper. Cell Res. 261, 91 -103 (2000) Schwartz and Baron, Curr. Opin. Cell Biol. 11, 197 -202 (1999) Vuori and Ruoslahti, Science 266, 1576 -1578 (1994)