NMR biomolecular studies

Multidimensional Nuclear Magnetic Resonance (NMR) studies have been recently emerged to better understand protein internal dynamics and determine protein structures. In these studies, a major time-consuming step is the generation of accurate NMR spectral peak lists, which usually requires significant amount of manual interaction. Human intervention becomes tedious and error prone at higher dimensions. Accurate peak lists are difficult to extract because of many ambiguities in low dimensional NMR spectra and because of a large number of low signal-to-noise ratio spectral peaks in high dimensional NMR spectra. Therefore, accurate spectral parameters estimation and adaptivity to contrasts between the sensitivity and resolution of multidimensional NMR spectrum require rigorous and computationally efficient analysis in a way which incorporates knowledge of the underlying physics. Moreover, multidimensional NMR experiments in protein structure and dynamics studies vary in complexity and specifications. The focus of my research in this field is to develop novel statistical procedures for analysis of multi-dimensional NMR spectrum, which are computational efficient, fully automated and applies to most of NMR experimental specifications.

Papers:
1. Serban, N. (2007), "MICE: Multiple-peak Identification, Estimation, and Characterization", Biometrics, 63,531-539.[.pdf ]
2. Serban, N. (submitted 2008), "Multi-Dimensional Biomolecular NMR Studies: Noise Reduction and Component Identification".[.pdf ]