Publications:Improving automatic peptide mass fingerprint protein identification by combining many peak sets
Property "Publisher" has a restricted application area and cannot be used as annotation property by a user. Property "Author" has a restricted application area and cannot be used as annotation property by a user. Property "Author" has a restricted application area and cannot be used as annotation property by a user. Property "Author" has a restricted application area and cannot be used as annotation property by a user. Property "Author" has a restricted application area and cannot be used as annotation property by a user. Property "Author" has a restricted application area and cannot be used as annotation property by a user. Property "Author" has a restricted application area and cannot be used as annotation property by a user.
| Title | Improving automatic peptide mass fingerprint protein identification by combining many peak sets |
|---|---|
| Author | |
| Year | 2004 |
| PublicationType | Journal Paper |
| Journal | Journal of chromatography. B |
| HostPublication | |
| Conference | |
| DOI | http://dx.doi.org/10.1016/j.jchromb.2004.04.010 |
| Diva url | http://hh.diva-portal.org/smash/record.jsf?searchId=1&pid=diva2:237403 |
| Abstract | An automated peak picking strategy is presented where several peak sets with different signal-to-noise levels are combined to form a more reliable statement on the protein identity. The strategy is compared against both manual peak picking and industry standard automated peak picking on a set of mass spectra obtained after tryptic in gel digestion of 2D-gel samples from human fetal fibroblasts. The set of spectra contain samples ranging from strong to weak spectra, and the proposed multiple-scale method is shown to be much better on weak spectra than the industry standard method and a human operator, and equal in performance to these on strong and medium strong spectra. It is also demonstrated that peak sets selected by a human operator display a considerable variability and that it is impossible to speak of a single “true” peak set for a given spectrum. The described multiple-scale strategy both avoids time-consuming parameter tuning and exceeds the human operator in protein identification efficiency. The strategy therefore promises reliable automated user-independent protein identification using peptide mass fingerprints. |
