Tandem mass spectrometry has emerged as a cornerstone of high throughput proteomic studies owing in part to various high throughput search engines which are used to interpret these tandem mass spectra. However, majority of experimental tandem mass spectra cannot be interpreted by any existing search engines or other methods. There are many reasons why this happens. However, one of the most important reasons is that majority of experimental spectra are of too poor quality to be interpretable. It wastes time to interpret these "uninterpretable" spectra by search engines or other methods. Therefore, if a powerful filter that could eliminate those spectra with poor quality is applied before any interpretations, it could significantly save the interpretation time of a whole set of spectra using search engines such as SEQUEST. This paper proposes a novel method to assess the quality of tandem mass spectra, and then use this method to develop a powerful filter that can eliminate majority of poor quality spectra while losing very minority of high quality spectra. First, a number of features are proposed to describe the quality of tandem mass spectra. The proposed method maps each tandem spectrum into a feature vector. Then Fisher linear discriminant analysis (FLDA) is employed to construct the classifier (the filter) which discriminates the high quality spectra from the poor quality ones. The proposed method has been tested on two tandem mass spectra datasets acquired by ion trap mass spectrometers. Computational experiments illustrate that the proposed method outperforms existing ones. The proposed method is generic, and is expected to be applicable to assessing the quality of spectra acquired by instruments other than ion trap mass spectrometers.
展开▼
