Aligning peaks across multiple mass spectrometry data sets using a scale-space based approach

We proposed a scale-space approach to automatically align multiple MS peak sets without manual parameter determination. It is more robust against noise than the hierarchical clustering method. In addition, it is possible to embed intensity information into the alignment framework, thus generalizing current approaches that use only the m/z information during the alignment of peaks. Our tests showed that this generalization brought some extra advantages for peak alignment, although we did not show concrete examples here due to the space limitation.