Abstract
The purpose of this study was to develop a method of classifying cancers to specific diagnostic categories based on their gene expression signatures using artificial neural networks (ANNs). We trained the ANNs using the small, round blue-cell tumors (SRBCTs) as a model. These cancers belong to four distinct diagnostic categories and often present diagnostic dilemmas in clinical practice. The ANNs correctly classified all samples and identified the genes most relevant to the classification. Expression of several of these genes has been reported in SRBCTs, but most have not been associated with these cancers. To test the ability of the trained ANN models to recognize SRBCTs, we analyzed additional blinded samples that were not previously used for the training procedure, and correctly classified them in all cases. This study demonstrates the potential applications of these methods for tumor diagnosis and the identification of candidate targets for therapy.
This is a preview of subscription content, access via your institution
Access options
Subscribe to this journal
Receive 12 print issues and online access
$209.00 per year
only $17.42 per issue
Buy this article
- Purchase on SpringerLink
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
Similar content being viewed by others
References
Pizzo, P.A. Principles and practice of pediatric oncology. (Lippincott-Raven, Philadelphia, 1997).
Triche, T.J. & Askin, F.B. Neuroblastoma and the differential diagnosis of small-, round-, blue- cell tumors. Hum. Pathol. 14, 569–595 (1983).
Taylor, C. et al. Diagnosis of Ewing's sarcoma and peripheral neuroectodermal tumour based on the detection of t(11;22) using fluorescence in situ hybridisation. Br. J. Cancer 67, 128–133 (1993).
McManus, A.P., Gusterson, B.A., Pinkerton, C.R. & Shipley, J.M. The molecular pathology of small round-cell tumours—relevance to diagnosis, prognosis, and classification. J. Pathol. 178, 116–121 (1996).
Khan, J. et al. Gene expression profiling of alveolar rhabdomyosarcoma with cDNA microarrays. Cancer Res. 58, 5009–5013 (1998).
Alizadeh, A.A. et al. Distinct types of diffuse large B-cell lymphoma identified by gene expression profiling. Nature 403, 503–511 (2000).
Bittner, M. et al. Molecular classification of cutaneous malignant melanoma by gene expression profiling. Nature 406, 536–540 (2000).
Golub, T.R. et al. Molecular classification of cancer: class discovery and class prediction by gene expression monitoring. Science 286, 531–537 (1999).
Bishop, C.M. Neural Networks for Pattern Recognition. (Clarendon Press, Oxford, 1995).
Heden, B., Ohlin, H., Rittner, R. & Edenbrandt, L. Acute myocardial infarction detected in the 12-lead ECG by artificial neural networks. Circulation 96, 1798–1802 (1997).
Silipo, R., Gori, M., Taddei, A., Varanini, M. & Marchesi, C. Classification of arrhythmic events in ambulatory electrocardiogram, using artificial neural networks. Comput. Biomed. Res. 28, 305–318 (1995).
Ashizawa, K. et al. Artificial neural networks in chest radiography: application to the differential diagnosis of interstitial lung disease. Acad. Radiol. 6, 2–9 (1999).
Abdolmaleki, P. et al. Neural network analysis of breast cancer from MRI findings. Radiat. Med. 15, 283–293 (1997).
Eisen, M.B., Spellman, P.T., Brown, P.O. & Botstein, D. Cluster analysis and display of genome-wide expression patterns. Proc. Natl. Acad. Sci. USA 95, 14863–14868 (1998).
deLapeyriere, O. et al. Expression of the Fgf6 gene is restricted to developing skeletal muscle in the mouse embryo. Development 118, 601–611 (1993).
Jaakkola, S. et al. Amplification of fgfr4 gene in human breast and gynecological cancers. Int. J. Cancer. 54, 378–382 (1993).
Shaoul, E., Reich-Slotky, R., Berman, B. & Ron, D. Fibroblast growth factor receptors display both common and distinct signaling pathways. Oncogene 10, 1553–1561 (1995).
Kovar, H. et al. Overexpression of the pseudoautosomal gene MIC2 in Ewing's sarcoma and peripheral primitive neuroectodermal tumor. Oncogene 5, 1067–1070 (1990).
Kumar, S., Perlman, E., Harris, C.A., Raffeld, M. & Tsokos, M. Myogenin is a specific marker for rhabdomyosarcoma: an immunohistochemical study in paraffin-embedded tissues. Mod. Pathol. 13, 988–993 (2000).
DeRisi, J. et al. Use of a cDNA microarray to analyse gene expression patterns in human cancer. Nature Genet. 14, 457–460 (1996).
Perou, C.M. et al. Molecular portraits of human breast tumours. Nature 406, 747–752 (2000).
Hedenfalk, I. et al. Gene-expression profiles in hereditary breast cancer. N. Engl. J. Med. 344, 539–548 (2001).
Brown, M.P. et al. Knowledge-based analysis of microarray gene expression data by using support vector machines. Proc. Natl. Acad. Sci. USA 97, 262–267 (2000).
Furey, T.S. et al. Support vector machine classification and validation of cancer tissue samples using microarray expression data. Bioinformatics 16, 906–914. (2000).
Mu, J. & Roach, P.J. Characterization of human glycogenin-2, a self-glucosylating initiator of liver glycogen metabolism. J. Biol. Chem. 273, 34850–34856 (1998).
Lee, M.G., Loomis, C. & Cowan, N.J. Sequence of an expressed human beta-tubulin gene containing ten Alu family members. Nucleic Acids Res. 12, 5823–5836 (1984).
Savchenko, V.L., McKanna, J.A., Nikonenko, I.R. & Skibo, G.G. Microglia and astrocytes in the adult rat brain: comparative immunocytochemical analysis demonstrates the efficacy of lipocortin 1 immunoreactivity. Neuroscience 96, 195–203 (2000).
Nagano, T. et al. Differentially expressed olfactomedin-related glycoproteins (Pancortins) in the brain. Brain Res. Mol. Brain Res. 53, 13–23 (1998).
Takahashi, Y., Campbell, E.A., Hirata, Y., Takayama, T. & Listowsky, I. A basis for differentiating among the multiple human Mu-glutathione S- transferases and molecular cloning of brain GSTM5. J. Biol. Chem. 268, 8893–8898 (1993).
Cavazzana, A.O., Miser, J.S., Jefferson, J. & Triche, T.J. Experimental evidence for a neural origin of Ewing's sarcoma of bone. Am. J. Pathol. 127, 507–518 (1987).
McKarney, L.A., Overall, M.L. & Dziadek, M. Myogenesis in cultures of uniparental mouse embryonic stem cells: differing patterns of expression of myogenic regulatory factors. Int. J. Dev. Biol. 41, 485–490 (1997).
Strohman, R.C., Micou-Eastwood, J., Glass, C.A. & Matsuda, R. Human fetal muscle and cultured myotubes derived from it contain a fetal-specific myosin light chain. Science 221, 955–957 (1983).
Song, W.K., Wang, W., Foster, R.F., Bielser, D.A. & Kaufman, S.J. H36-alpha 7 is a novel integrin alpha chain that is developmentally regulated during skeletal myogenesis. J. Cell. Biol. 117, 643–657 (1992).
Green, B.N. et al. Distinct expression patterns of insulin-like growth factor binding proteins 2 and 5 during fetal and postnatal development. Endocrinology 134, 954–962 (1994).
El-Badry, O.M. et al. Insulin-like growth factor II acts as an autocrine growth and motility factor in human rhabdomyosarcoma tumors. Cell Growth Differ. 1, 325–331 (1990).
Khan, J. et al. cDNA microarrays detect activation of a myogenic transcription program by the PAX3-FKHR fusion oncogene. Proc. Natl. Acad. Sci. USA 96, 13264–13269 (1999).
Holtrich, U., Brauninger, A., Strebhardt, K. & Rubsamen-Waigmann, H. Two additional protein-tyrosine kinases expressed in human lung: fourth member of the fibroblast growth factor receptor family and an intracellular protein-tyrosine kinase. Proc. Natl. Acad. Sci. USA 88, 10411–10415 (1991).
Hughes, S.E. Differential expression of the fibroblast growth factor receptor (FGFR) multigene family in normal human adult tissues. J. Histochem. Cytochem. 45, 1005–1019 (1997).
Chen, Y., Dougherty, E.R. & Bittner, M.L. Ratio-based decisions and the quantitative analysis of cDNA microarray images. Biomedical Optics 2, 364–374 (1997).
Jollife, I.T. Principal Component Analysis. (Springer, New York, 1986).
Acknowledgements
We thank K. Gayton, C. Tsokos, T. Fadiran, J. Lueders and R. Walker for their technical assistance; M. Ohlsson for valuable discussions on ANNs; R. Simon, M. Bittner, Y. Chen and S. Gruvberger for their helpful comments regarding the data analysis; and M. Tsokos, L. Helman and C. Thiele for cell lines supplied from the NCI. J.S.W. was in part supported by the Charles & Dana Nearburg Foundation. M.R. was in part supported by the Swedish Research Council and the Knut and Alice Wallenberg Foundation through the SWEGENE consortium. C.P. was in part supported by the Swedish Foundation for Strategic Research.
Author information
Authors and Affiliations
Corresponding authors
Supplementary information
Rights and permissions
About this article
Cite this article
Khan, J., Wei, J., Ringnér, M. et al. Classification and diagnostic prediction of cancers using gene expression profiling and artificial neural networks. Nat Med 7, 673–679 (2001). https://doi.org/10.1038/89044
Received:
Accepted:
Issue Date:
DOI: https://doi.org/10.1038/89044
This article is cited by
-
What are the Optimal Systemic Treatment Options for Rhabdomyosarcoma?
Current Treatment Options in Oncology (2024)
-
Novel gene signatures predicting and immune infiltration analysis in Parkinson’s disease: based on combining random forest with artificial neural network
Neurological Sciences (2024)
-
Fibroblast growth factor receptor type 4 as a potential therapeutic target in clear cell renal cell carcinoma
BMC Cancer (2023)
-
CD276-CAR T cells and Dual-CAR T cells targeting CD276/FGFR4 promote rhabdomyosarcoma clearance in orthotopic mouse models
Journal of Experimental & Clinical Cancer Research (2023)
-
Multimodal feature selection from microarray data based on Dempster–Shafer evidence fusion
The Journal of Supercomputing (2023)