Wavelet transform analysis of NMR structure ensembles to reveal internal fluctuations of enzymes

Abstract

Internal motions and flexibility are essential for biological functions in proteins. To assess the internal fluctuations and conformational flexibility of proteins, reliable computational methods are needed. In this study, wavelet transformation was used to filter out the noise and facilitate investigating the internal positional fluctuations of enzymes within nuclear magnetic resonance (NMR) structure ensembles. Moreover, potential active sites were identified by combining with positional fluctuation score, sequence conservation, and solvent accessible surface area. Among the total 107 catalytic residues in 44 examined enzymes, 69 residues were identified correctly. Our results suggest that wavelet transform analysis of structure ensemble is applicable to extract essential fluctuation information of proteins; furthermore, analysis of positional fluctuations is helpful for the identification of catalytic residues.

WaVPeak: picking NMR peaks through wavelet-based smoothing and volume-based filtering

“…Interchanges between these fields during the last 20 years have led to many new wavelet applications, especially image processing and de-noising noisy data. Wavelets have also been employed in various tasks to do with NMR signal processing ( Barache et al , 1997 ; Dancea and Güntert, 2005 ; Gronwald and Kalbitzer, 2004 ; Günther et al , 2000 , 2002 ; Hu et al , 2011 ; Lang et al , 1996 ; Neue, 1996 ; Shao et al , 2000 ). Such tasks include analyzing the dynamical behavior of NMR signals ( Barache et al , 1997 ; Hu et al , 2011 ; Neue, 1996 ), de-noising the NMR spectra ( Dancea and Güntert, 2005 ; Günther et al , 2000 ), suppressing water peaks from the spectra ( Gronwald and Kalbitzer, 2004 ; Günther et al , 2002 ), and increasing the resolution of the spectra ( Shao et al , 2000 ).…”

Filter design for molecular factor computing using wavelet functions

“…Wavelet transform (WT) has been successfully applied for chemical signal processing [20][21][22][23][24]. It has been shown that a signal can be represented by a very small number of wavelet functions, giving an efficient expression due to the orthogonality of wavelet functions.…”

Median Modified Wiener Filter for nonlinear adaptive spatial denoising of protein NMR multidimensional spectra

“…In the last two decades, computational community has played more and more important roles to simplify and accelerate this tedious structure determination process 2 3 4 5 6 7 8 9 . Peak picking has been treated as a signal processing problem and has been tackled by a variety of methods 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 . Resonance assignment, on the other hand, is often formulated as a graph-based problem to find the best mapping between spin systems and residues of the target protein.…”