VisualCV

The main research themes throughout my PhD revolved around protein structures, their dynamics and folding mechanisms. More specifically, my research focused on the dynamics and folding of proteins found within ubiquitin signalling pathways, concentrating on comparative studies of molecular dynamics (MD) simulations with experimental data for two protein domains, p62-UBA and NBR1-PB1. This was achieved using both all-atom and united-atom force fields with implicit and explicit descriptions of the solvent. Following these comparative equilibrium studies, the unfolding of these two domains were studied using umbrella sampling and Φ-restrained MD simulations. As part of my PhD training, I have experience in basic molecular biology having performed protein expression for production of unlabelled and labelled samples for NMR spectroscopy to obtain comparative data for my simulations.

Temperature gradients within an NMR sample can generate convection currents. These currents contribute to phase errors and signal loss in the pulsed field gradient spin echo spectra. An example are the spectra used for diffusion-ordered spectroscopy (DOSY) where such distortion of the spectra leads to an overestimation of the apparent diffusion coefficients. This is particularly noticeable when taking diffusion measurements away from ambient temperature; in mobile solvents such measurements may prove difficult or impossible. The convective motion of spins is coherent, whilst that from diffusion is random. The effects of convection and diffusion can thus be separated by refocusing the dephasing of spins caused by convection. The measured signal attenuation is then solely a consequence of diffusion and relaxation. When I started my MPhil several pulse sequences based on this principle had been described that compensated to varying degrees for the effects of convection; I developed new pulse sequences that offered significantly improved NMR performance while retaining convection compensation. (See Abstracts) 

Gradient shimming methods are now widely used and another example of where convection can have deleterious effects. Typically, these methods employ deuterium spin or gradient echoes to map the magnetic field within the sample, giving rapid and efficient convergence to optimum field homogeneity. Gradient shimming methods are frequently thought to be incompatible with variable temperature operation, although gradient shimming is often successful for modest departures from room temperature. The limiting factor is the loss of echo signal as a result of sample convection, which originates from temperature gradients within the sample. The magnitudes of such gradients depend upon a number of factors, notably probe design, air flow rate, sample viscosity, and operating temperature. I realised and developed with Gareth Morris a new method for variable temperature gradient shimming (see publications); a simple modification was made to the existing gradient shimming pulse sequences that compensated for the effects of convection, resulting in good performance throughout the liquid range of common solvents.