S1: Modeling drug induced liver injury

Modeling drug induced liver injury using co-expression networks

S1: Modeling drug induced liver injury, OpenTox USA 2017

Dr. James Stevens 


Lilly Research Laboratories


Research Fellow


Animals and man are complex adaptive systems having emergent properties.  These emergent properties are fundamental for normal biological function but present challenges in modeling and predicting responses to environmental perturbations including toxic drugs and chemicals.  Considerable effort has been expended modeling responses to specific classes of chemical using in vivo and in vitro models as a strategy to underpin risk assessment.  However, modeling the components of the underlying biological systems can also reveal important information and allow responses to chemical exposure to be considered both at the level of unique responses that depend on the chemical agent (e.g. metabolism to a reactive intermediate) or as inherent responses of the biological system itself (e.g. regeneration and repair).  The presentation will describe how co-expression algorithms can be used to model liver gene expression data as co-expressed networks or modules with each module representing the aggregate function of all genes in the network.  The networks are then used to model the biological responses of liver to drug-induced injury.