Cluster-based Peak Alignment Techniques for LC-MS Data Analysis
| Autor: | Hui-Yin Chang, 張彙音 |
|---|---|
| Rok vydání: | 2009 |
| Druh dokumentu: | 學位論文 ; thesis |
| Popis: | 97 Identifying proteomics markers to classify diseases by using mass spectrometry with high-performance liquid chromatography (LC-MS) has been a trend recently. However, since the experimental errors, one traditional problem (also known as the peak-shifting problem) occurred in the preprocessing of mutltiple LC-MS data analysis is that the identical peptides from multiple samples may have different retention time drifts. In our study, we proposed an algorithm, namely PeakAlign, to solve the peak-shifting problem. Our algorithm consists of two phases, the adjustment phase and the alignment phase. In the adjustment phase, the LOESS regression method is used to calculate the different shifting values among peaks along the retention time. In the alignment phase, a novel cluster-based technique based on the distance constraint is applied to align the adjusted peaks. To evaluate the PeakAlign, we used two real LC-MS datasets as well as a set of generated semi-synthetic datasets to evaluate the accuracy and similarity of the alignment results. The experimental results show that the performance of our algorithm is much better than that of other methods, such as the DTW and the SlidingWin algorithms. |
| Databáze: | Networked Digital Library of Theses & Dissertations |
| Externí odkaz: |
|