Abstract of Zhang's talk
From PublicWiki
Revision as of 21:01, 11 January 2008; Simonlin (Talk | contribs)
(diff) ←Older revision | Current revision | Newer revision→ (diff)
(diff) ←Older revision | Current revision | Newer revision→ (diff)
Monday, January 14, 2008
9:30 am – 10:30 am Conference Room 1206 676 N St. Clair St Development of peak detection and alignment algorithms for label-free quantitative proteomics by LC-MS Yonghong Zhang, PhD Windber Research Institute, Windber, PA Mass spectrometry is widely used in biomedical research for protein identification and quantitative analysis. Coupling of HPLC and MS (LC- MS) provides a quantitative means for protein analysis with a much higher resolution and through-put than traditional methods such as 2D gel. Quantitative proteomics using LC-MS is a promising method for biomarker discovery and disease profiling. However, the data of LC-MS analyses can not be readily compared because neither features (peaks) representing protein peptides are available, nor features representing the same protein peptide across different samples are matched. Peak detection and alignment play a critical role in quantitative proteomics using LC-MS and can dramatically affect downstream data analysis. Here we present novel methods to detect peptide peaks in LC- MS analyses as a 3D feature (m/z, retention time, and intensity) without data smoothing, and to align peaks across different samples by direct comparison of m/z and retention time matching by linear regression. Performance of the methods is evaluated and compared with methods implemented in MZmine. The result shows that our method is more effective and reliable. Hosts: Warren Kibbe and Simon Lin
