Computational modeling of biased signaling in G protein-coupled receptors

Summary

The complexity of pharmacological efficacy at G protein-Coupled Receptors opens up the possibility of selective control of their signaling pathways, i.e., a ligand can selectively activate alternative signal transduction pathways to those activated by the native ligand at the same receptor. This potentially offers great opportunities in many fields of biology, including mammalian reproduction. Kinetic experiments that measure the activity of several effectors downstream of a receptor after ligand binding with respect to time, are now widely available. The proliferation of crosstalk, redundancy and multiple regulatory mechanisms within signaling cascades prevent use of a direct statistical approach to understand and infer ligand efficacy. This thesis project seeks to explore how Boolean network methodology can help to compare different ligands while considering the complexity of signaling pathways.

However, the sparsity of available data means that a family of Boolean networks must be learned instead of a single network. To reduce the number of networks, we need to obtain more experimental data, which is usually costly. An alternative is to identify conditions that help to reduce the number of inferred networks computationally without performing more experiments.

Our objectives are to :

i) develop a tool using logic programming (answer set programming) to infer networks which will be able to predict time series signals,

ii) sample networks from a large set of networks,

iii) identify a way to characterize different dynamical behaviors within a set of networks, and iv) develop an algorithm based on the entropy measure to select informative conditions to make this set of dynamic behaviors less variable.

Contact :

• Pamela Romero Jofré