Please use this identifier to cite or link to this item:
http://hdl.handle.net/1942/16501
Title: | Comparison of the Mahalanobis Distance and Pearson's χ(2) Statistic as Measures of Similarity of Isotope Patterns. | Authors: | ZAMANZAD GHAVIDEL, Fatemeh CLAESEN, Jurgen BURZYKOWSKI, Tomasz VALKENBORG, Dirk |
Issue Date: | 2014 | Source: | JOURNAL OF THE AMERICAN SOCIETY FOR MASS SPECTROMETRY, 25 (2), p. 293-296 | Abstract: | To extract a genuine peptide signal from a mass spectrum, an observed series of peaks at a particular mass can be compared with the isotope distribution expected for a peptide of that mass. To decide whether the observed series of peaks is similar to the isotope distribution, a similarity measure is needed. In this short communication, we investigate whether the Mahalanobis distance could be an alternative measure for the commonly employed Pearson’s χ2 statistic. We evaluate the performance of the two measures by using a controlled MALDI-TOF experiment. The results indicate that Pearson’s χ2 statistic has better discriminatory performance than the Mahalanobis distance and is a more robust measure. | Keywords: | similarity statistics; isotope distributions; mass spectral data interpretation; bioinformatics; mahalanobis distance | Document URI: | http://hdl.handle.net/1942/16501 | ISSN: | 1044-0305 | e-ISSN: | 1879-1123 | DOI: | 10.1007/s13361-013-0773-z | ISI #: | 000330828700015 | Rights: | © American Society for Mass Spectrometry, 2013. | Category: | A1 | Type: | Journal Contribution | Validations: | ecoom 2015 |
Appears in Collections: | Research publications |
Files in This Item:
File | Description | Size | Format | |
---|---|---|---|---|
art%3A10.1007%2Fs13361-013-0773-z.pdf Restricted Access | 267.02 kB | Adobe PDF | View/Open Request a copy |
SCOPUSTM
Citations
3
checked on Sep 3, 2020
WEB OF SCIENCETM
Citations
3
checked on Oct 13, 2024
Page view(s)
132
checked on Apr 17, 2023
Download(s)
106
checked on Apr 17, 2023
Google ScholarTM
Check
Altmetric
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.