Computational Investigation into the Search Mechanism of a DNA Glycosylase DNA Damage Recognition by AlkD

Project Description

Using molecular dynamics-based research techniques, we probe for atomistic insight into the mechanisms of nucleic acid interacting enzymes. Currently, AlkD, a non-canonical DNA glycosylase, is being investigated. In performing hydrolysis of alkylated nucleotides, AlkD is the first known glycosylase to keep the nucleotides in the DNA duplex and not initially perform a base-flipping step. This observation makes AlkD ideal for microsecond-timescale molecular dynamics studies, in which crystal structures from Eichmann et al are run in a dichotomous procedure -- the methylated DNA in AlkD and an artificial non-methylated DNA state are simulated. The analysis of this data is achieved using programming languages such as Fortran, and scripting languages such as Python, and TCL scripts in VMD. As such, we are able to parse these theoretical representations of protein-DNA dynamics and distill from them a story regarding, in this case, the way AlkD is able to located damaged nucleotides in a sea of DNA duplexes. Additionally, we apply alchemical simulation methods such as Thermodynamic Integration and Umbrella Sampling, in which we apply artificial biases to specific coordinates, such as the positions of certain atoms, to restrain these systems in particular states -- and through these methods determine thermodynamic properties such as the free energy required to unbend the DNA from its preferred structure within AlkD.


  • Molecular Dynamics
  • Helical Parameter Analysis
  • Thermodynamic Integration
  • Enhanced Sampling Techniques
  • Metadynamics
  • Umbrella Sampling