A Systematic Low-Level Analysis for Both SELDI- and MALDI-TF-MS Data with Improved Peak Alignment Algorithm

• Activity Number: 365
• Type: Contributed
• Date/Time: Wednesday, August 6, 2008 : 8:30 AM to 10:20 AM
• Sponsor: Biometrics Section
• Author(s): Xuelian Wei*+ and Kerchau Li+ and James LeBlanc and Thomas Drake
• Companies: University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles and University of California, Los Angeles
• Keywords: SELDI-TF-MS ; MALDI-TF-MS ; peak alignment ; low-level analysis ; mass spectrometry

Abstract

In this study, we first summarized the pre-process steps for mass spectrometry analysis proposed recently and provided a systematical pre-process procedure suitable for both SELDI- and MALDI-TF-MS data. And then we focus on the peak alignment step and proposed so-called correlation-based hierarchical clustering (CBHC). This new method is motivated form the complete linkage hierarchical cluster (CLHC), but we add many important modifications to best ultimate all the information from both the locations and the shapes of detected peaks to form a better peak alignment. We validate our method on two real mass spectrometry datasets. We show that CBHC produces more reasonable peak alignment than CLHC and the protein profiles detected by CBHC have lower variations among replicated spectra and help to increase the sensitivity of follow-up biomarker detection comparing with CLHC and Ciphergen.