He Lin, Han Xi, Ma Bin
David R. Cheriton School of Computer Science, University of Waterloo, Waterloo, Ontario, Canada N2L 3G1, Canada.
J Bioinform Comput Biol. 2013 Aug;11(4):1350007. doi: 10.1142/S0219720013500078. Epub 2013 Apr 11.
De novo sequencing derives the peptide sequence from a tandem mass spectrum without the assistance of protein databases. This analysis has been indispensable for the identification of novel or modified peptides in a biological sample. Currently, the speed of de novo sequencing algorithms is not heavily affected by the number of post-translational modification (PTM) types in consideration. However, the accuracy of the algorithms can be degraded due to the increased search space. Most peptides in a proteomics research contain only a small number of PTMs per peptide, yet the types of PTMs can come from a large number of choices. Therefore, it is desirable to include a large number of PTM types in a de novo sequencing algorithm, yet to limit the number of PTM occurrences in each peptide to increase the accuracy. In this paper, we present an efficient de novo sequencing algorithm, DeNovoPTM, for such a purpose. The implemented software is downloadable from http://www.cs.uwaterloo.ca/~l22he/denovo_ptm .
从头测序在没有蛋白质数据库辅助的情况下，从串联质谱中推导肽序列。这种分析对于鉴定生物样品中的新型或修饰肽来说必不可少。目前，从头测序算法的速度受所考虑的翻译后修饰（PTM）类型数量的影响不大。然而，由于搜索空间增加，算法的准确性可能会下降。蛋白质组学研究中的大多数肽每个肽仅包含少量的PTM，但PTM的类型可以有大量选择。因此，希望在从头测序算法中纳入大量的PTM类型，但要限制每个肽中PTM出现的数量以提高准确性。在本文中，我们提出了一种高效的从头测序算法DeNovoPTM，以实现这一目的。所实现的软件可从http://www.cs.uwaterloo.ca/~l22he/denovo_ptm下载。
