Identifying Interactions Controlling Protein Structure And Transition Kinetics Via Efficient Simulation Methods
Advances in hardware and molecular force fields have given a boost to computational studies of the thermodynamics and dynamics of transitions involving small protein systems. Because these studies are still hampered by the vast amount of CPU time required, new approaches and sampling optimization strategies are still needed. In this work, several schemes were developed and used to increase the simulation efficiency of various proteins and give insights on their structure, kinetics and mechanism. Our studies focus on the recognition of particular markers that assist in antibody design or lead to protein misfolding and aggregation. Concerning structural identification, the application of novel techniques based on the Replica Exchange Method is illustrated by a mutagenesis analysis seeking to "humanize" the hypervariable regions of a llama heavy-chain antibody. Regarding kinetic and mechanistic characterization, the application of optimization schemes of the Forward Flux Sampling Method is demonstrated by the study of structural transitions for the Alanine Dipeptide and the Tryptophan Cage synthetic protein. Our results are in good agreement with previous experimental studies performed on these systems and further characterize the pathways and "transition states" traversed in the studied transitions.
Dissertation or Thesis