**5. Conclusions**

The results presented here show that very good classification results can be obtained from DT, MCA, SVM, and LDA classifiers, even if the dataset contains no information. Studies using any of these methods should carefully examine whether the results are due to some underlying biology or are just fortuitous. Performing comparable examinations on randomly generated feature values, or performing analysis of the same data after the

A Comparison of Biomarker and Fingerprint-Based Classifiers of Disease 199

Guyon, I., Weston, J., & Barnill, S. (2002) Gene Selection for Cancer Classification Using Support Vector Machines. *Machine Learning*, Vol. 46, No. 1, pp. 389-422 Ho, D.W., Yang, Z.F., Wong, B.Y., Kwong, D.L., Sham, J.S., Wei, W.I., et al. (2006). Surface-

Kohonen, T. (1988). Self-Organizing and Associative Memory, 2nd ed., New York, Springer-

LeCun, Y., Denker. J., & Solla, S. (1990). Optimal Brain Damage, In *Advances In Neural Information Processing System II*, Morgan Kaufmann, San Mateo, CA, USA Liu, W., Guan, M., Wu, D., Zhang, Y., Wu, Z., Xu, M., & Lu, Y. (2005). Using Tree

Luke, B.T., & Collins, J.R. (2008). *Examining the Significance of Fingerprint-based Classifiers.*

Noble, W. (2004). Support Vector Machine Applications in Computational Biology, In:

Ornstein, D.K., Rayford, W., Fusaro, V.A., Conrads, T.P., Ross, S.J., Hitt, B.A., et al. (2004).

Petricoin, E.F., Fishman, D.A., Conrads, T.P., Veenstra, T.D., & Liotta, L.A. (2003). Proteomic

Petricoin, E.F., Rajapaske, V., Herman, E.H., Arekani, A.M., Ross, S., Johann, D., et al. (2004).

Ransohoff, D.F. (2005a). Lessons From Controversy: Ovarian Cancer Screening and Serum Proteomics. *Journal of the National Cancer Institute,* Vol. 97, pp. 315-319 Ransohoff, D.F. (2005b). Bias as a Threat to the Validity of Cancer Molecular-marker

Srinivasan, R., Daniels, J., Fusaro, V., Lundqvist, A., Killian, J.K., Geho, D., et al. (2006).

Stone, J.H., Rajapakse, V.N., Hoffman, G.S., Specks, U., Merkel, P.A., Spiera, R.F., et al.

Tang, Z., Han, L., Lin, H., et al. (2007). Derivation of Stable Microarray Cancer

BMC Bioinformatics, Vol. 9, No. 1, (December 17), p. 545

15.0 ng/ml. *Journal of Urology*, Vol. 172, pp. 1302-1305

Research. *Nature Reviews Cancer*, Vol. 5, pp. 142-149

Pattern Analysis. *Experimental Hematology,* Vol. 34, pp. 796-801

granulomatosis. *Arthritis & Rheumatism,* Vol. 52, pp. 902-10

*BMC Bioinformatics,* Vol. 4, p. 24/comments

107

Verlag

USA

Vol. 47, pp. 456-62

32, Suppl. 1, pp. 122-30

9996-10003

Enhanced Laser Desorption/Ionization Time-of-Flight Mass Spectrometry Serum Protein Profiling to Identify Nasopharyngeal Carcinoma. *Cancer,* Vol. 107, pp. 99-

Analysis Pattern and SELDI-TOF-MS to Discriminate Transitional Cell Carcinoma of the Bladder Cancer From Noncancer Patients. *European Urology,*

*Kernel Methods in Computational Biology*, pp. 71-92, MIT Press*,* Cambridge, Mass,

Serum Proteomic Profiling Can Discriminate Prostate Cancer From Benign Prostates In Men With Total Prostate Specific Antigen Levels Between 2.5 and

Pattern Diagnostics: Producers and Consumers in the era of correlative science.

Toxicoproteomics: Serum Proteomic Pattern Diagnostics For Early Detection of Drug Induced Cardiac Toxicities and Cardioprotection. *Toxicologic Pathology,* Vol.

Accurate Diagnosis of Acute Graft-versus-host Disease Using Serum Proteomic

(2005). A serum proteomic approach to gauging the state of remission in Wegener's

Differentiating Signatures Using Consensus Scoring of Multiple Random Sampling and Gene-ranking Consistency Evaluation. *Cancer Research,* Vol. 67, No., 20, pp.

group labels of the samples has been scrambled, should be a necessary part of these investigations.

In contract, a biomarker-based classifier such as BMDK should perform poorly if the dataset contains no biological information. Even when reasonably good results are obtained, the putative biomarkers used in the classifier should be carefully examined to ensure that they really distinguish samples in one group from the other.

### **6. Acknowledgement**

The authors would like to thank Dr. Jih-H Chen who constructed the recursive SVM and LDA programs and generated their results. This work was funded in whole with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E and the Center for Biomedical Informatics and Information Technology (CBIIT)/Cancer Biomedical Informatics Grid (caBIG) ISRCE yellow task #09-260 to NCI-Frederick. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the United States Government.

### **7. References**


group labels of the samples has been scrambled, should be a necessary part of these

In contract, a biomarker-based classifier such as BMDK should perform poorly if the dataset contains no biological information. Even when reasonably good results are obtained, the putative biomarkers used in the classifier should be carefully examined to ensure that they

The authors would like to thank Dr. Jih-H Chen who constructed the recursive SVM and LDA programs and generated their results. This work was funded in whole with federal funds from the National Cancer Institute, National Institutes of Health, under Contract No. HHSN261200800001E and the Center for Biomedical Informatics and Information Technology (CBIIT)/Cancer Biomedical Informatics Grid (caBIG) ISRCE yellow task #09-260 to NCI-Frederick. The content of this publication does not necessarily reflect the views or policies of the Department of Health and Human Services, nor does mention of trade names, commercial products, or organizations imply endorsement by the United States

Bellman, R.E. & Rand Corporation (1957). *Dynamic Programming,* Princeton University

Boser. B., Guyon, I. &Vapnik, V. (1992). A Training Algorithm For Optimal Margin

Browers, F.M., Petricoin, E.F. 3rd, Ksinantova, L., Breza, J., Rajapakse, V., Ross, S.,

Conrads, T.P., Fusaro, V.A., Ross, S., Johann, D., Rajapakse, V., Hitt, B.A., et al. (2004). High-

Dudoit, S., Fridlyand, J., & Speed, T.P. (2002). Comparison of Discrimination Methods For

Fukunaga, K. (1990). *Introduction to Statistical Pattern Recognition*, Academic Press, 2nd

Golub, G. & Van Loan, C. (1983). *Matrix Computations*. The John Hopkins University Press,

Classifiers, In: *Proceedings of the 5th Annual Workshop on Computational Learning*

et al. (2005). Low Molecular Weight Proteomic Information Distinguishes Metastatic From Benign Pheochromocytoma. *Endocrine-Related Cancer* Vol. 12, pp.

Resolution Serum Proteomic Features For Ovarian Cancer Detection. *Endocrine-*

Classification of Tumors Using Gene Expression Data. *Journal of the American* 

Bellman, R.E. (1961). *Adaptive Control Processes: A Guided Tour,* Princeton University Press

Bellman, R.E. (2003). *Dynamic Programming,* Courier Dover Publications

Breiman, L. (2001). Random Forests. *Machine Learning*, Vol. 45, pp. 5-32

*Statistical Association,* Vol. 97, No. 457, pp. 77-87

*Theory*, pp. 114-152, New York, NY, USA

*Related Cancer,* Vol. 11, pp. 163-78

Edition, Boston, Mass, USA

Baltimore, USA

really distinguish samples in one group from the other.

investigations.

Government.

**7. References** 

Press

263-72

**6. Acknowledgement** 


**1. Introduction** 

**10** 

*Japan* 

). RNS

**8-Nitroguanine, a Potential Biomarker** 

**Inflammation-Related Carcinogenesis** 

Recently, chronic inflammation induced by infection has been postulated to be an important risk factor of various cancers (Schetter et al., 2010; Aggarwal & Sung, 2011; Kamp et al., 2011; Rook & Dalgleish, 2011; Trinchieri, 2011). Many malignancies arise from areas of infection and inflammation (Balkwill & Mantovani, 2001; Coussens & Werb, 2002). Epidemiological and experimental studies have provided evidence showing that chronic infection and inflammation contribute to a substantial part of environmental carcinogenesis (Coussens & Werb, 2002; IARC, 2003). It has been estimated that chronic inflammation accounts for approximately 25 % of human cancers (Hussain S. P. & Harris, 2007). International Agency for Research on Cancer (IARC) has estimated that infectious diseases account for approximately 18 % of cancer cases worldwide (IARC, 2003). During inflammation, nitric oxide (NO) and reactive oxygen species (ROS) are generated from inflammatory cells and considered to play the key role in carcinogenes (Hofseth et al., 2003a; Hofseth et al., 2003b; Hussain S. P. et al., 2003; Ohshima et al., 2003). Inducible nitric oxide synthase (iNOS) catalyzes the production of NO particularly during inflammation, leading to generation of

various reactive nitrogen species (RNS), such as NOx and peroxynitrite (ONOO-

generated during infection with influenza viruses can mediate the formation of 8 nitroguanine, a nitrative lesion of nucleic acids, via ONOO- formation (Maeda H. & Akaike, 1998; Akaike et al., 2003). 8-Nitroguanine formed in DNA is chemically unstable, and thus can be spontaneously released, resulting in the formation of an apurinic site (Yermilov et al., 1995a). The apurinic site can form a pair with adenine during DNA synthesis, leading to G:C-to-T:A transversions (Kawanishi & Hiraku, 2006) (Fig. 1). Thus, 8-nitroguanine is a potentially mutagenic DNA lesion, which can participate in initiation and promotion in the

**to Evaluate the Risk of** 

Shiho Ohnishi2, Raynoo Thanan2,

*2Faculty of Pharmaceutical Sciences,* 

Yusuke Hiraku3 and Shosuke Kawanishi2

*Suzuka University of Medical Science, Suzuka, Mie,* 

*Suzuka University of Medical Science, Suzuka, Mie, 3Department of Environmental and Molecular Medicine, Mie University Graduate School of Medicine, Tsu, Mie,* 

Ning Ma1, Mariko Murata3,

*1Faculty of Health Science,* 

Vapnik, V.N. (1998). *Statistical Learning Theory*, John Wiley & Sons, New York, NY, USA

