Feature extraction in the analysis of proteomic mass spectra

Article (PDF Available)inPROTEOMICS 6(7):2095-100 · July 2006with22 Reads
DOI: 10.1002/pmic.200500459 · Source: PubMed
Feature extraction or biomarker selection is a critical step in disease diagnosis and knowledge discovery based on protein MS. Many studies have discussed the classification methods applied in proteomics; however, few could be found to address feature extraction in detail. In this paper, we developed a systematic approach for the extraction of mass spectrum peak apex and peak area with special emphasis on noise filtration and peak calibration. Application to a head and neck cancer data generated at the Eastern Virginia Medical School [Wadsworth, J. T., Somers, K. D., Cazares, L. H., Malik, G. et al.., Clin. Cancer Res. 2004, 10, 1625-1632] revealed that the new feature extraction method would yield consistent and highly discriminatory biomarkers.
Proteomics 2006, 6, 4203 4203
Comparative proteomic analysis of Helicobacter
pylori strains associated with iron deficiency anemia
by S. A. Park et al., vol. 6, issue 4, pp. 1319–1328.
DOI 10.1002/pmic.200500293
The list of authors of this paper should read:
Shin Ae Park
, HyangWoo Lee
, Myung Hee Hong
, Young
Wook Choi
, Yon Ho Choe
, Bo Young Ahn
1, 3
, Yang Je Cho
Dong Su Kim
4, 5
and Na Gyong Lee
Department of Bioscience and Biotechnology, Institute of
Bioscience, Sejong University, Seoul, Korea
Department of Pediatrics, Samsung Medical Center, Sung-
kyunkwan University, Seoul, Korea
R&D Center, EyeGene Inc., Seoul, Korea
Research Division, Genomine Inc., Pohang, Kyungbuk,
Division of Molecular and Life Sciences and Systems Bio-
Dynamics Research Center, POSTECH, Pohang, Kyungbuk,
790–784, Korea
Feature extraction in the analysis of proteomic mass
by X. Wang et al., vol. 6, issue 7, pp. 2095–2100.
DOI 10.1002/pmic.200500459
Please note that Equation 3 on page 2096 of this article
should read:
(x) I
(x, w)+k6s(x, w)
Glutamine regulates the expression of proteins with a
potential health-promoting effect in human intestinal
Caco-2 cells
by K. Lenaerts et al., vol. 6, issue 8, pp. 2454–2464.
DOI 10.1002/pmic.200500692
Please note that in Table 1 on page 2459 the Swiss-Prot
accession numbers for spots 1 and 2 should read Q15293 and
P61978 respectively.
Proteomic analysis of a meningococcal outer
membrane vesicle vaccine prepared from the
group B strain NZ98/254
by C. Vipond et al., vol. 6, issue 11, pp. 3400–3413.
DOI 10.1002/pmic.200500821
The list of authors of this paper should read:
Caroline Vipond
, Janet Suker
, Christopher Jones
, Chris-
toph Tang
, Ian M. Feavers
and Jun X. Wheeler
Department of Bacteriology, National Institute for Biologi-
cal Standards and Control, South Mimms, Hertfordshire,
Laboratory of Molecular Structure, National Institute for
Biological Standards and Control, South Mimms, Hertford-
shire, UK
CMMI, Flowers Building, Imperial College, South Ken-
sington, UK
Proteomic analysis of factors released from
p21-overexpressing tumour cells
by C. A. Currid et al., vol. 6, issue 13, pp. 3739–3753.
DOI 10.1002/pmic.200500787
The last line of the abstract was supposed to read:
“may contribute to the observed mitogenic and anti-apopto-
tic paracrine activity of p21-expressing cells
and NOT p22-expressing cells as it says.
© 2006 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim www.proteomics-journal.com
    • "Subsequently, peak picking was performed by finding all the local maxima and eliminating those with intensities lower than a non-uniform threshold proportional to the noise level (Currie, 1999) (Yasui et al., 2003). Since the mass spectra could be inaccurately aligned after the calibration procedure, a maximum tolerance distance equal to 600 ppm of the m/z value was accepted for the comparison (Fushiki et al., 2006) (Wang et al., 2006). Finally, a classification of peaks based on the peak detection rate (PDR) was performed, which was expressed by the ratio between the number of spectra containing the considered peak and the total number of analysed spectra (Mantini et al., 2007). "
    [Show abstract] [Hide abstract] ABSTRACT: The growth of spoiling yeasts in beverages results in reduced quality, economic and image losses. Therefore, biochemical and DNA-based identification methods have been developed but are mostly time-consuming and laborious. Matrix-Assisted-Laser-Desorption/Ionization-Time-Of-Flight Mass Spectrometry (MALDI-TOF MS) could deliver discriminative peptide mass fingerprints within minutes and could thus be a rapid and reliable tool for identification and differentiation. However, routine analysis of yeasts by MALDI-TOF MS is yet impaired by low reproducibility and effects of different physiological states of organisms on the reliability of the identification method are still controversial. The aim of this study was to optimize sample preparation and measurement parameterization using three spoilage yeasts (Saccharomyces cerevisiae var. diastaticus, Wickerhamomyces anomalus and Debaryomyces hansenii). The influence of environmental or physiological parameters including oxygen availability, different nutrients, cell density and growth phase were analysed and revealed small differences in mass fingerprints. Yeasts grown in the presence or absence of oxygen were precisely differentiated along these differences in mass fingerprints and a crude classification of growth phase was possible. Cell concentration did not affect the spectra distinctly, neither qualitatively nor quantitatively, and an influence of available nutrients could not be measured in each case. However, core mass peaks remained constant under all tested conditions enabling reliable identification.
    Article · Dec 2013
    • "abundance, such as the peptide peak intensity (height or area of a peak), the peptide precursor ion peak height, and the peak height of product ions, can be extracted. Using such information individually or combinatorially, numerous label-free methods have been developed, including two extensively applied but fundamentally different strategies: quantitation based on spectral counting [10] and peptide ion peak area [11]. Spectral counting estimates protein abundance by counting the number of spectra matched to peptides from a speci�c protein. "
    [Show abstract] [Hide abstract] ABSTRACT: To address the challenges associated with differential expression proteomics, label-free mass spectrometric protein quantification methods have been developed as alternatives to array-based, gel-based, and stable isotope tag or label-based approaches. In this paper, we focus on the issues associated with label-free methods that rely on quantitation based on peptide ion peak area measurement. These issues include chromatographic alignment, peptide qualification for quantitation, and normalization. In addressing these issues, we present various approaches, assembled in a recently developed label-free quantitative mass spectrometry platform, that overcome these difficulties and enable comprehensive, accurate, and reproducible protein quantitation in highly complex protein mixtures from experiments with many sample groups. As examples of the utility of this approach, we present a variety of cases where the platform was applied successfully to assess differential protein expression or abundance in body fluids, in vitro nanotoxicology models, tissue proteomics in genetic knock-in mice, and cell membrane proteomics.
    Full-text · Article · Jan 2013
    • "The peak-intensity regions generation step proved efficient in boosting up the performance of the system under the Majority Vote combination rule. Previous studies [23,29] have utilized similar classifier combination schemes, but from a different perspective. According to these studies, the predictions of individual classifier members of the combination schemes were based on coding aggregated information extracted from each MS-spectrum. "
    [Show abstract] [Hide abstract] ABSTRACT: In this study, a pattern recognition system is presented for improving the classification accuracy of MS-spectra by means of gathering information from different MS-spectra intensity regions using a majority vote ensemble combination. The method starts by automatically breaking down all MS-spectra into common intensity regions. Subsequently, the most informative features (m/z values), which might constitute potential significant biomarkers, are extracted from each common intensity region over all the MS-spectra and, finally, normal from ovarian cancer MS-spectra are discriminated using a multi-classifier scheme, with members the Support Vector Machine, the Probabilistic Neural Network and the k-Nearest Neighbour classifiers. Clinical material was obtained from the publicly available ovarian proteomic dataset (8-7-02). To ensure robust and reliable estimates, the proposed pattern recognition system was evaluated using an external cross-validation process. The average overall performance of the system in discriminating normal from cancer ovarian MS-spectra was 97.18% with 98.52% mean sensitivity and 94.84% mean specificity values.
    Full-text · Article · Dec 2009
Show more

Recommended publications

Discover more