86 / 161
Conformational Ensembles from Experimental Data
and Computer Simulations
Poster Abstracts
82
47-POS
Board 7
Molecular and Metadynamics to Predict Activation in Experimentally Validated Kinase
Domain Mutations
E. Joseph Jordan
1
, Ravi Radhakrishnan
1,2
.
1
The University of Pennsylvania, Phildelphia, PA, USA,
2
The University of Pennsylvania,
Phildelphia, PA, USA.
Kinase proteins are important in cell growth and division and are frequently mutated in cancer.
Numerous studies have shown a good correlation between increased catalytic efficiency in
mutant kinases and increased ability of mutant cell lines to grow in cell culture. Kinases have
distinct active and inactive conformations and this makes them amenable to simulations to
predict whether mutations bias the conformation to the active state, which should correlate with
increased catalytic efficiency. In this study, we run molecular dynamics simulations of a series of
experimentally validated mutants from BRAF, HER2, and ALK kinases and compare the
dynamics to those of wild type simulations as assessed by both molecular and metadynamics.
This allows us to make predictions of the effect of mutations on kinase activation with greater
than 60% accuracy using simple metrics like changes in hydrogen bond occupancy during the
course of a simulation. We believe that this method could eventually be used in guiding
personalized medical treatments as both the number of observed mutations and the number of
available drugs increases going forward.