Improved peak detection and quantification of mass spectrometry data acquired from surface-enhanced laser desorption and ionization by denoising spectra with the undecimated discrete wavelet transform

ASIA unversity > 醫學暨健康學院 > 生物科技學系 > 期刊論文 > Item 310904400/4616

Please use this identifier to cite or link to this item: http://asiair.asia.edu.tw/ir/handle/310904400/4616

Title:	Improved peak detection and quantification of mass spectrometry data acquired from surface-enhanced laser desorption and ionization by denoising spectra with the undecimated discrete wavelet transform
Authors:	Coombes, K.R.;Tsavachidis, S.;Morris, J. S.;Baggerly, K. A.;Mien-Chie Hung;Kuerer, H. M.
Date:	2005-01
Issue Date:	2009-11-27 05:57:05 (UTC+0)
Publisher:	Asia University
Abstract:	Background: Mass spectrometry, especially surface enhanced laser desorption and ionization (SELDI) is increasingly being used to find disease-related proteomic patterns in complex mixtures of proteins derived from tissue samples or from easily obtained biological fluids such as serum, urine, or nipple aspirate fluid. Questions have been raised about the reproducibility and reliability of peak quantifications using this technology. For example, Yasui and colleagues opted to replace continuous measures of the size of a peak by a simple binary indicator of its presence or absence in their analysis of a set of spectra from prostate cancer patients. Methods: We collected nipple aspirate fluid from breast cancer patients and from healthy women. Samples were pooled to form a single quality control (QC) sample, separated into aliquots, and stored. We extracted protein from an aliquot of the QC sample and hybridized it to two spots on each of three SELDI ProteinChip arrays using weak cation exchange (WCX2; Ciphergen). The experiment was repeated on four successive days. We developed a novel algorithm for low level processing of SELDI spectra, including denoising with the undecimated discrete wavelet transform (UDWT), baseline correction, peak detection and quantification. We evaluated this algorithm for consistency and reproducibility across the 24 replicate spectra. Results: The UDWT provides a computationally efficient method for decomposing mass spectra into a noise component and a signal component consisting of peaks and baseline. The noise levels are consistent and mostly uncorrelated across spectra. The subsequent baseline correction step yields spectra consisting of isolated peaks or peak clusters separated by flat regions. Our method detects more peaks than the method implemented in the Ciphergen software, and the peaks it detects are reproducibly found across replicate spectra. After normalization to the total ion current and log transformation, the mean coefficient of variation of the peak heights is 10.6%. Software to implement these methods is freely available. Conclusions: The method for low-level processing of SELDI spectra described here provides substantial improvements over existing methods. In particular, denoising spectra using the UDWT appears to be an important step toward obtaining more accurate results. It both improves the reproducibility of the peak quantifications and supplies tools that will make it possible to investigate the variations in the technology more carefully.
Relation:	PROTEOMICS 5(16):4107-4117
Appears in Collections:	[生物科技學系] 期刊論文

Files in This Item:

File	Description	Size	Format
		0Kb	Unknown	649	View/Open
310904400-4616.doc		34Kb	Microsoft Word	476	View/Open

Loading...