Skip to main content

Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles and JavaScript.

  • Protocol
  • Published:

Preprocessing, classification modeling and feature selection using flow injection electrospray mass spectrometry metabolite fingerprint data

Nature Protocols volume 3, pages 446–470 (2008)Cite this article

Abstract

Metabolome analysis by flow injection electrospray mass spectrometry (FIE-MS) fingerprinting generates measurements relating to large numbers of m/z signals. Such data sets often exhibit high variance with a paucity of replicates, thus providing a challenge for data mining. We describe data preprocessing and modeling methods that have proved reliable in projects involving samples from a range of organisms. The protocols interact with software resources specifically for metabolomics provided in a Web-accessible data analysis package FIEmspro (http://users.aber.ac.uk/jhd) written in the R environment and requiring a moderate knowledge of R command-line usage. Specific emphasis is placed on describing the outcome of modeling experiments using FIE-MS data that require further preprocessing to improve quality. The salient features of both poor and robust (i.e., highly generalizable) multivariate models are outlined together with advice on validating classifiers and avoiding false discovery when seeking explanatory variables.

This is a preview of subscription content, access via your institution

Access options

Buy this article

  • Purchase on Springer Link
  • Instant access to full article PDF
Buy now

Prices may be subject to local taxes which are calculated during checkout

Figure 1: Workflow in FIE-MS data analysis.
Figure 2: TIC checking for potential outlying samples and batch effects in FIE-MS data using FIEmspro function ticstats.
Figure 3: Use of baseline correction to improve FIE-MS fingerprint data representing pathogen-challenged B. distachyon plants.
Figure 4: Evaluation of an optimum number of k with the FIEmspro function koptimp.
Figure 5: Effect of log transformation on the signal variance to signal intensity dependency within FIE fingerprints representing the metabolome of B. distachyon challenged with the rice blast fungus.
Figure 6: Effect of TIC normalization on fingerprint data representing B. distachyon leaves after either 3 or 4 d infection with rice blast fungus.
Figure 7: Outlier detection in FIE-MS fingerprint data representing B. distachyon leaves after 3 d infection with rice blast fungus using PCA (pccomp) and FIEmspro function outl.det.
Figure 8: PCA of FIE-MS fingerprints representing a time course of B. distachyon infected with the rice blast fungus.
Figure 9: LDA and HCA of FIE-MS representing disease progression in B. distachyon plants infected with the rice blast fungus.
Figure 10: Examples of RF statistics derived from the classification of FIE-MS fingerprint data representing B. distachyon plants during a time course of infection with rice blast fungus.
Figure 11: Validation of discrimination models using FIE-MS fingerprint data that describe disease progression in B. distachyon infected with the rice blast fungus.
Figure 12: Relationship between RF importance score and variable ranking for explanatory value in binary comparisons of B. distachyon leaves taken at different times after infection with the rice blast fungus (ST, a significance threshold for explanatory variables).

Similar content being viewed by others

References

  1. Somorjai, R.L., Dolenko, B. & Baumgartner, R. Class prediction and discovery using gene microarray and proteomics mass spectroscopy data: curses, caveats, cautions. Bioinformatics 19, 1484–1491 (2003).

    Article  CAS  Google Scholar 

  2. Berrar, D., Bradbury, I. & Dubitzky, W. Avoiding model selection bias in small-sample genomic datasets. Bioinformatics 22, 1245–50 (2006).

    Article  CAS  Google Scholar 

  3. BragaNeto, U.M. & Dougherty, E.R. Is cross-validation valid for small-sample microarray classification? Bioinformatics 20, 374–380 (2004).

    Article  CAS  Google Scholar 

  4. Lyons-Weiler, J. et al. Assessing the statistical significance of the achieved classification error of classifiers constructed using serum peptide profiles, and a prescription for random sampling repeated studies for massive high-throughput genomic and proteomic studies. Cancer Inform. 1, 53–77 (2005).

    CAS  PubMed  Google Scholar 

  5. Broadhurst, D.I. & Kell, D.B. Statistical strategies for avoiding false discoveries in metabolomics and related experiments. Metabolomics 2, 171–196 (2006).

    Article  CAS  Google Scholar 

  6. Saghatelian, A. & Cravatt, B.F. Global strategies to integrate the proteome and metabolome. Curr. Opin. Chem. Biol. 9, 62–68 (2005).

    Article  CAS  Google Scholar 

  7. EinDor, L., Zuk, O. & Domany, E. Thousands of samples are needed to generate a robust gene list for predicting outcome in cancer. Proc Natl. Acad. Sci. USA 103, 5923–5928 (2006).

    Article  CAS  Google Scholar 

  8. Dyaz-Uriarte, R. Supervised methods with genomic data: a review and cautionary view. Data Analysis and Visualization in Genomics and Proteomics. pp 193–214 Wiley, New York, (2005).

    Chapter  Google Scholar 

  9. Fawcett, T. ROC Graphs: Notes and Practical Considerations for Data Mining Researchers. Technical report HPL-2003-4. HP Laboratories, Palo Alto, CA, Available at http://www.hpl.hp.com/techreports/2003/HPL-2003-4.pdf (2003).

    Google Scholar 

  10. Mukherjee, S., Roberts, S.J. & van der Laan, M.J. Data-adaptive test statistics for microarray data. Bioinformatics 21, 108–114 (2005).

    Article  Google Scholar 

  11. Sima, C. & Dougherty, E.R. What should be expected from feature selection in small-sample settings. Bioinformatics 22, 2430–2436 (2006).

    Article  CAS  Google Scholar 

  12. Enot, D.P., Beckmann, M., Overy, D. & Draper, J. Predicting interpretability of metabolome models based on behavior, putative identity, and biological relevance of explanatory signals. Proc. Natl. Acad. Sci. USA 103, 14865–14870 (2006).

    Article  CAS  Google Scholar 

  13. Kell, D.B., Darby, R.M. & Draper, J. Genomic computing. Explanatory analysis of plant expression profiling data using machine learning. Plant Physiol. 126, 943–951 (2001).

    Article  CAS  Google Scholar 

  14. Catchpole, G.S. et al. Hierarchical metabolomics demonstrates substantial compositional similarity between genetically modified and conventional potato crops. Proc. Natl. Acad. Sci. USA 102, 14458–14462 (2005).

    Article  CAS  Google Scholar 

  15. Goodacre, R., Vaidyanathan, S., Dunn, W.B., Harrigan, G.G. & Kell, D.B. Metabolomics by numbers: acquiring and understanding global metabolite data. Trends Biotechnol. 22, 245–252 (2004).

    Article  CAS  Google Scholar 

  16. Bino, R.J. et al. Potential of metabolomics as a functional genomics tool. Trends Plant Sci. 9, 418–425 (2004).

    Article  CAS  Google Scholar 

  17. Fiehn, O. et al. Metabolite profiling for plant functional genomics. Nat. Biotechnol. 18, 1157–1161 (2000).

    Article  CAS  Google Scholar 

  18. Sumner, L.W., Mendes, P. & Dixon, R.A. Plant metabolomics: large-scale phytochemistry in the functional genomics era. Phytochemistry 62, 817–836 (2003).

    Article  CAS  Google Scholar 

  19. Nicholson, J.K. & Wilson, I.D. Understanding 'global' systems biology: metabonomics and the continuum of metabolism. Nat. Rev. Drug Discov. 2, 668–676 (2003).

    Article  CAS  Google Scholar 

  20. Roessner, U., Wagner, C., Kopka, J., Trethewey, R.N. & Willmitzer, L. Simultaneous analysis of metabolites in potato tuber by gas chromatography-mass spectrometry. Plant J 23, 131–142 (2000).

    Article  CAS  Google Scholar 

  21. Tolstikov, V.V. & Fiehn, O. Analysis of highly polar compounds of plant origin: Combination of hydrophilic interaction chromatography and electrospray ion trap mass spectrometry. Anal. Biochem. 301, 298–307 (2002).

    Article  CAS  Google Scholar 

  22. Beckmann, M., Enot, D.P., Overy, D.P. & Draper, J. Representation, comparison and interpretation of metabolome fingerprint data for total composition analysis and quality trait investigation in potato cultivars. J. Agricultural and Food Chemistry 55, 3444–3451 (2007).

    Article  CAS  Google Scholar 

  23. Dear, G.J., James, A.D. & Sarda, S. Ultra-performance liquid chromatography coupled to linear ion trap mass spectrometry for the identification of drug metabolites in biological samples. Rapid Commun. Mass Spectrom. 20, 1351–1360 (2006).

    Article  CAS  Google Scholar 

  24. Wagner, C., Sefkow, M. & Kopka, J. Construction and application of a mass spectral and retention time index database generated from plant GC/EI-TOF-MS metabolite profiles. Phytochemistry 62, 887–900 (2003).

    Article  CAS  Google Scholar 

  25. Jonsson, P. et al. A strategy for identifying differences in large series of metabolomic samples analyzed by GC/MS. Anal. Chem. 76, 1738–1745 (2004).

    Article  CAS  Google Scholar 

  26. Vorst, O. et al. A non-directed approach to the differential analysis of multiple LC–MS-derived metabolic profiles. Metabolomics 1, 169–180 (2005).

    Article  CAS  Google Scholar 

  27. Ward, J.L., Harris, C., Lewis, J. & Beale, M.H. Assessment of H-1 NMR spectroscopy and multivariate analysis as a technique for metabolite fingerprinting of Arabidopsis thaliana. Phytochemistry 62, 949–957 (2003).

    Article  CAS  Google Scholar 

  28. Allen, J. et al. High-throughput classification of yeast mutants for functional genomics using metabolic footprinting. Nat. Biotechnol. 21, 692–696 (2003).

    Article  CAS  Google Scholar 

  29. Scholz, M., Gatzek, S., Sterling, A., Fiehn, O. & Selbig, J. Metabolite fingerprinting: detecting biological features by independent component analysis. Bioinformatics 20, 2447–2454 (2004).

    Article  CAS  Google Scholar 

  30. Aharoni, A. et al. Nontargeted metabolome analysis by use of Fourier Transform Ion Cyclotron Mass Spectrometry. OMICS 6, 217–234 (2002).

    Article  CAS  Google Scholar 

  31. Smedsgaard, J. & Frisvad, J.C. Using direct electrospray mass spectrometry in taxonomy and secondary metabolite profiling of crude fungal extracts. J Microbiol. Methods 25, 5–17 (1996).

    Article  CAS  Google Scholar 

  32. Dunn, W.B., Bailey, N.J. & Johnson, H.E. Measuring the metabolome: current analytical technologies. Analyst 130, 606–625 (2005).

    Article  CAS  Google Scholar 

  33. Beckmann, M., Parker, D., Enot, D.P., Duval, E. & Draper, J. High-throughput, nontargeted metabolite fingerprinting using nominal mass flow injection electrospray mass spectrometry. Nat. Protoc. 3, 486–504 (2008).

    Article  CAS  Google Scholar 

  34. Overy, D.P. et al. Explanatory signal interpretation and metabolite identification strategies for nominal mass FIE-MS metabolite fingerprints. Nat. Protoc. 3, 471–485 (2008).

    Article  CAS  Google Scholar 

  35. Parker, D. et al. Rice blast infection of Brachypodium distachyon as a model system to study dynamic host/pathogen interactions. Nat. Protoc. 3, 435–445 (2008).

    Article  CAS  Google Scholar 

  36. Enot, D.P., Beckmann, M. & Draper, J. Detecting a difference—assessing generalisability when modelling metabolome fingerprint data in longer term studies of genetically modified plants. Metabolomics 3, 335–347 (2007).

    Article  Google Scholar 

  37. Enot, D.P. & Draper, J. Statistical measures for testing substantial equivalence of GM plant genotypes in a multivariate context. Metabolomics 3, 349–355 (2007).

    Article  CAS  Google Scholar 

  38. Jain, A.K., Murty, M.N. & Flynn, P.J. Data clustering: a review. ACM Computing Surveys (CSUR) 31, 264–323 (1999).

    Article  Google Scholar 

  39. Manly, B.F.J. Multivariate Statistical Methods: A Primer. Chapman & Hall/CRC, London (2004).

    Book  Google Scholar 

  40. Zhang, C., Lu, X. & Zhang, X. Significance of gene ranking for classification of microarray samples. EEE/ACM Transactions on Computational Biology and Bioinformatics 3, 312–320 (2006).

    Google Scholar 

  41. Ransohoff, D.F. Rules of evidence for cancer molecular-marker discovery and validation. Nat. Rev. Cancer 4, 309–313 (2004).

    Article  CAS  Google Scholar 

  42. Davis, C.A. et al. Reliable gene signatures for microarray classification: assessment of stability and performance. Bioinformatics 22, 2356–2363 (2006).

    Article  CAS  Google Scholar 

  43. Wu, B. et al. Comparison of statistical methods for classification of ovarian cancer using mass spectrometry data. Bioinformatics 19, 1636–1643 (2003).

    Article  CAS  Google Scholar 

  44. Cristianini, N. & Shawe-Taylor, J. An Introduction to Support Vector Machines and Other Kernel-Based Learning Methods. Cambridge University Press, Cambridge (2000).

    Book  Google Scholar 

  45. Zhu, C., Kitagawa, H. & Faloutsos, C. Example-based outlier detection for high dimensional datasets. IPSJ Digital Courier 1, 234–243 (2005).

    Article  Google Scholar 

  46. Craig, A., Cloarec, O., Holmes, E., Nicholson, J.K. & Lindon, J.C. Scaling and normalization effects in NMR spectroscopic metabonomic data sets. Anal. Chem. 78, 2262–2267 (2006).

    Article  CAS  Google Scholar 

  47. Hastie, T., Tibshirani, R. & Friedman, J. The Elements of Statistical Learning: Data Mining, Inference, and Prediction. Springer Series in Statistics, New York (2001).

    Book  Google Scholar 

  48. Good, P. Permutation Tests: A Practical Guide to Resampling Methods for Testing Hypotheses. Springer Series in Statistics, Heidelberg (2000).

    Book  Google Scholar 

  49. Efron, B. Estimating the error rate of a prediction rule: improvement on cross-validation. J. Am. Stat. Assoc. 78, 316–331 (1983).

    Article  Google Scholar 

  50. Sing, T., Sander, O., Beerenwinkel, N. & Lengauer, T. ROCR: visualizing classifier performance in R. Bioinformatics 21, 3940–3941 (2005).

    Article  CAS  Google Scholar 

  51. Fu, W.J., Carroll, R.J. & Wang, S. Estimating misclassification error with small samples via bootstrap cross-validation. Bioinformatics 21, 1979–1986 (2005).

    Article  CAS  Google Scholar 

  52. Thomaz, C.E. et al. Using a maximum uncertainty LDA-based approach to classify and analyse MR brain images. Lecture Notes in Computer Science: Medical Image Computing and Computer-Assisted Intervention—MICCAI 2004, pp 291–3 Springer, Berlin, 291–300 (2004).

    Google Scholar 

  53. Yang, J. & Yang, J. Why can LDA be performed in PCA transformed space? Pattern Recognition 36, 563–566 (2003).

    Article  Google Scholar 

  54. Breiman, L. Random Forests. Machine Learning 45, 5–32 (2001).

    Article  Google Scholar 

  55. Zar, J.H. Biostatistics. 2nd edn. (Prentice-Hall, Englewood Cliffs, New Jersey, 1984).

    Google Scholar 

  56. Dietterich, T.G. Ensemble methods in machine learning. Lecture Notes in Computer Science 1857, 1–15 (2000).

    Google Scholar 

  57. Vaidyanathan, S., Kell, D.B. & Goodacre, R. Flow-injection electrospray ionization mass spectrometry of crude cell extracts for high-throughput bacterial identification. J. Am. Soc. Mass Spectrom. 13, 118–128 (2002).

    Article  CAS  Google Scholar 

  58. Roessner, U. & Luedemann, A. et al. Metabolic profiling allows comprehensive phenotyping of genetically or environmentally modified plant systems. Plant Cell 13, 11–29 (2001).

    Article  CAS  Google Scholar 

  59. Mazzella, N. et al. Use of electrospray ionization mass spectrometry for profiling of crude oil effects on the phospholipid molecular species of two marine bacteria. Rapid Commun. Mass Spectrom. 19, 3579–3588 (2005).

    Article  CAS  Google Scholar 

  60. Favretto, D., Piovan, A., Filippini, R. & Caniato, R. Monitoring the production yields of vincristine and vinblastine in Catharanthus roseus from somatic embryogenesis. Semiquantitative determination by flow-injection electrospray ionization mass spectrometry. Rapid Commun. Mass Spectrom. 15, 364–369 (2001).

    Article  CAS  Google Scholar 

  61. Rashed, M.S., Al-Ahaidib, L.Y., Aboul-Enein, H.Y., Al-Amoudi, M. & Jacob, M. Determination of L-pipecolic acid in plasma using chiral liquid chromatography-electrospray tandem mass spectrometry. Clin. Chem. 47, 2124–2130 (2001).

    CAS  PubMed  Google Scholar 

  62. Overy, S.A. et al. Application of metabolite profiling to the identification of traits in a population of tomato introgression lines. J. Exp. Bot. 56, 287–296 (2005).

    Article  CAS  Google Scholar 

  63. Goodacre, R., York, E.V., Heald, J.K. & Scott, I.M. Chemometric discrimination of unfractionated plant extracts analyzed by electrospray mass spectrometry. Phytochemistry 62, 859–863 (2003).

    Article  CAS  Google Scholar 

  64. Koulman, A. et al. High-throughput direct-infusion ion trap mass spectrometry: a new method for metabolomics. Rapid Commun. Mass Spectrom. 21, 421–428 (2007).

    Article  CAS  Google Scholar 

  65. Martinez, A.M. & Kak, A.C. PCA versus LDA. IEEE Transactions on: Pattern Analysis and Machine Intelligence 23, 228–233 (2001).

    Google Scholar 

  66. Windeatt, T. Vote counting measures for ensemble classifiers. Pattern Recognition 36, 2743–2756 (2003).

    Article  Google Scholar 

  67. R_Development_Core_Team. R. A Language and Environment for Statistical Computing. R Foundation for Statistical Computing. Vienna, Austria, ISBN 3-900051-900007-900050, URL http://www.R-project.org (2006).

  68. Benjamini, Y. & Hochberg, Y. Controlling the false discovery rate: a practical and powerful approach to multiple testing. J. R. Stat. Soc. Ser. B Stat. Methodol. 57, 289–300 (1995).

    Google Scholar 

  69. Storey, J.D. A direct approach to false discovery rates. J. R. Stat. Soc. Ser. B Stat. Methodol. 64, 479–498 (2002).

    Article  Google Scholar 

Download references

Acknowledgements

Financial support was provided for W.L., M.B. and D.P.O by the UK Food Standards Agency G03012 programme. D.P.E. and D.P. were funded by grants MET20483 and BBD0069531 respectively, from the UK Biotechnology and Biological Sciences Research Council.

Author information

Author notes

  1. David P Overy

    Present address: Present address: CIBE, Merck Sharp & Dohme de España, Madrid 28027, Spain.,

Authors and Affiliations

  1. Institute of Biological Sciences, Aberystwyth University, Aberystwyth, SY23 3DA, UK

    David P Enot, Wanchang Lin, Manfred Beckmann, David Parker, David P Overy & John Draper

Authors
  1. David P Enot
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Wanchang Lin
    View author publications

    You can also search for this author in PubMed Google Scholar

  3. Manfred Beckmann
    View author publications

    You can also search for this author in PubMed Google Scholar

  4. David Parker
    View author publications

    You can also search for this author in PubMed Google Scholar

  5. David P Overy
    View author publications

    You can also search for this author in PubMed Google Scholar

  6. John Draper
    View author publications

    You can also search for this author in PubMed Google Scholar

Corresponding author

Correspondence to John Draper.

Rights and permissions

Reprints and permissions

About this article

Cite this article

Enot, D., Lin, W., Beckmann, M. et al. Preprocessing, classification modeling and feature selection using flow injection electrospray mass spectrometry metabolite fingerprint data. Nat Protoc 3, 446–470 (2008). https://doi.org/10.1038/nprot.2007.511

Download citation

  • Published:

  • Issue Date:

  • DOI: https://doi.org/10.1038/nprot.2007.511

This article is cited by

Comments

By submitting a comment you agree to abide by our Terms and Community Guidelines. If you find something abusive or that does not comply with our terms or guidelines please flag it as inappropriate.

Search

Advanced search

Quick links

Nature Briefing

Sign up for the Nature Briefing newsletter — what matters in science, free to your inbox daily.

Get the most important science stories of the day, free in your inbox. Sign up for Nature Briefing