Drug Target Identification Via Computational Modeling
We use computational models of biological pathways in disease to estimate the range of possible responses to various hypothetical drugs, given patient variability and limited pathway-relevant data.
The first step in drug development is to identify the target biological molecule that the drug will disable or enhance. Choosing a target often relies on intuition, implicit assumptions, and unrealistic experimental models of the disease. Computational models of biological pathways are an underutilized way to validate conclusions from intuition and experimental models. In this project, we use ordinary differential equation models of biochemical kinetics to evaluate and compare hypothetical drug targets. Currently we are studying targets in the complement system, a set of immune system proteins in the blood, for treating sepsis and rare inflammatory kidney diseases. For each target, we estimate the response of the disease condition to various drug amounts and drug interaction strengths. For each estimate, we also calculate lower and upper bounds that quantify the uncertainty and variability arising from insufficient experimental data and differences among patients. For those estimated responses with wide bounds, we use the model to compare methods for narrowing the bounds, such as collecting additional data or restricting treatment to patient subpopulations with a consistent and positive drug response. Our methodology is generalizable to other diseases and provides a useful tool for decision making in early-phase drug development.