As the genetic content is internally located within DNA duplexed form, it has long been hypothesized that DNA undergoes a series of thermally induced conformational changes that assist in protein recognition events. The biological mechanisms for protein-DNA interactions have long been sought after, as little is still known mechanistically about how these complexes form. To study the local contributions to these breathing modes several atomistic simulations of DNA oligonucleotides were generated and analyzed by statistical models to predict metastable conformational states, the system timescales, and the kinetic pathways between states.
In order to sample time-series DNA constructs, microsecond molecular dynamics (MD) simulation were performed. MD simulations provide atomstic resolution of macromolecules in explicit solvent and with modern computational workflows can extend well into microsecond timescales. While MD is a powerful tool, it creates a tremendous amount of time-dependent data. In recent years, Markov State Models (MSM), which project the dynamics of MD simulations onto discrete coordinates that follow a Markov chain, have become an invaluable tool to model and describe the kinetics of these large datasets. These models can be coarse-grained for chemical insight, however there does not yet exist a method which consistently and ``crisply'' describe the metastable barriers.
To address this, I developed a new method, called Gradient Adaptive Decomposition (GRAD), which optimizes the macrostate model by refining borders with respect to the gradient along the free energy surface. The proposed method requires only a small number of initial microstates because it corrects for errors produced by limited number of seeds. Whereas many methods rely on fuzzy, or overlapping, partitions for proper statstical analysis of timescales, GRAD retains accuracy and crisp decomposition.
I present a workflow of GRAD refined MSM to analyze the long timescale MD simulations of DNA oligonucleotides to assess the stacking conformational dynamics of DNA. Evaluating the complex network of transitions accessible found evidence suggesting that chiral directed mechanisms are critical in how DNA bases unstack. I explore how these local effects may be significant to long timescale dynamics and the biological impact in relation to breathing fluctuations.
This dissertation includes unpublished co-authored material.
| Identifer | oai:union.ndltd.org:uoregon.edu/oai:scholarsbank.uoregon.edu:1794/23126 |
| Date | 10 April 2018 |
| Creators | Romano, Pablo |
| Contributors | Guenza, Marina |
| Publisher | University of Oregon |
| Source Sets | University of Oregon |
| Language | en_US |
| Detected Language | English |
| Type | Electronic Thesis or Dissertation |
| Rights | All Rights Reserved. |
Page generated in 0.0015 seconds