Inference of dynamic biological networks based on responses to drug perturbations

Area of Science:

• Computational biology
• Systems biology
• Cancer research

Background:

• Targeted protein inhibitors are crucial in cancer research.
• Current models infer static pathway behavior from drug perturbation data.
• Existing methods capture only a steady-state snapshot, limiting dynamic insights.

Purpose of the Study:

• To extend modeling frameworks for drug sensitivity prediction and combination therapy design.
• To infer dynamic models that can generate observed static target inhibition maps.
• To explore deterministic and stochastic approaches for modeling biological system dynamics.

Main Methods:

• Inferred Boolean networks from binarized drug sensitivities under target inhibition.
• Investigated Markov chain models from stochastic perspectives.
• Analyzed the inverse problem of dynamic models generating static target inhibition maps.

Main Results:

• Identified deterministic Boolean networks capable of producing observed drug sensitivities.
• Developed stochastic Markov chain models consistent with target inhibition data.
• Extended static target inhibition maps to dynamic pathway models.

Conclusions:

• Dynamic modeling offers deeper insights into drug responses than static models.
• This framework enhances prediction of drug sensitivity and combination therapy efficacy.
• The approach provides a more comprehensive understanding of cancer signaling pathways.