On de novo interpretation of tandem mass spectra for peptide identification
peer-reviewed · Proceedings of the seventh annual international conference on Research in computational molecular biology · 2003
| Date | 2003-04-10 |
| Type | peer-reviewed |
| Venue | Proceedings of the seventh annual international conference on Research in computational molecular biology |
| Publisher | ACM |
| Contribution | algorithm |
| DOI | 10.1145/640075.640077 |
| Citations (OpenAlex) | 59 |
Abstract
The correct interpretation of tandem mass spectra is a difficult problem, even when it is limited to scoring peptides against a database. De novo sequencing is considerably harder, but critical when sequence databases are incomplete or not available. In this paper we build upon earlier work due to Dancik et al., and Chen et al. to provide a dynamic programming algorithm for interpreting de novo spectra. Our method can handle most of the commonly occurring ions, including a; b; y, and their neutral losses. Additionally, we shift the emphasis away from sequencing to assigning ion types to peaks. In particular, we introduce the notion of core interpretations, which allow us to give confidence values to individual peak assignments, even in the absence of a strong interpretation. Finally, we introduce a systematic approach to evaluating de novo algorithms as a function of spectral quality. We show that our algorithm, in particular the core-interpretation, is robust in the presence of measurement error, and low fragmentation probability.
Methods and tools
- Core interpretations (Bafna-Edwards): A dynamic programming de novo algorithm that assigns ion types (a, b, y and neutral losses) to peaks and introduces core interpretations to give confidence values to parts of a sequence.