The ability to move, motor control, is an adaptive behavior that emerges from the interactions of bone, muscle cells, motoneurons, sensory neurons, and the many other types of neurons which process sensory and motor information.  In order to improve our understanding of motor control, it is necessary to capture the complexity and diversity of the morphological and behavioral organization of all of these cells.  Computational models are an effective means of both organizing the large amounts of disparate information from experiments and of expanding our knowledge via the generation and testing of hypotheses.

Currently, I am working towards the goal of automated generation of multi-scale computational models of populations of morphologically accurate neurons.  I have developed software which can import morphological reconstructions of a population of neurons, perform a wide variety of morphometric analyses, and compare different populations.  Using the analysis as inputs, the software can then algorithmically generate populations of morphologically realistic "virtual" neurons. 

My long-term goal is to combine computational neuroscience with biomechanical modeling to generate neuro-musculo-skeletal models.  I believe that movement is the primary function of the nervous system.  By combining biophysically accurate sensory inputs (e.g. muscle tension information from group Ia afferents) derived from a musculoskeletal model in a physics engine with a cellular-level model of the spinal cord (beginning with motoneurons and monosynaptic reflexes), and then connecting the motoneuronal output to the musculoskeletal model will create a virtuous cycle which will greatly improve our understanding of sensorimotor integration and the nervous system as a whole.