Exhausted Jackknife Validation Exemplified by Prediction of Temperature Optimum in Enzymatic Reaction of Cellulases
- 100 Downloads
This was the continuation of our previous study along the same line with more focus on technical details because the data are usually divided into two datasets, one for model development and the other for model validation during the development of predictive model. The widely used validation method is the delete-1 jackknife validation. However, no systematical studies were conducted to determine whether the jackknife validation with different deletions works better because the number of validations with different deletions increases in a factorial fashion. Therefore it is only small dataset that can be used for such an exhausted study. Cellulase is an enzyme playing an important role in modern industry, and many parameters related to cellulase in enzymatic reactions were poorly documented. With increased interests in cellulases in bio-fuel industry, the prediction of parameters in enzymatic reactions is listed on agenda. In this study, two aims were defined (a) which amino acid property works better to predict the temperature optimum and (b) with which deletion the jackknife validation works. The results showed that the amino acid distribution probability works better in predicting the optimum temperature of catalytic reaction by cellulase, and the delete-4, more precisely one-fifth deletion, jackknife validation works better.
KeywordsCellulase Enzyme Jackknife validation Prediction Temperature optimum
This study was partly supported by Guangxi Science Foundation (07-109-001-3, 0907016, 10-046-06, 11-031-11, 2010GXNSFF013003, and 2010GXNSFA013046). The authors wish to thank the Library of Guangxi Zhuang Autonomous Region for purchasing the book, Biometry.
- 3.Levitin, A. (2003). Introduction to the design and analysis of algorithms (1st ed.). NJ: Pearson Education.Google Scholar
- 5.Enzyme Structures Database. (2011). http://www.ebi.ac.uk/thornton-srv/databases/enzymes/.
- 6.IntEnz. (2011). http://www.ebi.ac.uk/intenz/.
- 7.Comprehensive Enzyme Information System BRENDA. (2011). http://www.brenda-enzymes.info/php/result_flat.php4?ecno=220.127.116.11.
- 16.Kang, H. J., & Ishikawa, K. (2007). Journal of Microbiology and Biotechnology, 17, 1249–1253.Google Scholar
- 26.Cooper, G. M. (2004). The cell: A molecular approach (p. 51). Washington: ASM Press.Google Scholar
- 28.Chou, P. Y., & Fasman, G. D. (1978). Advances in Enzymology and Related Subjects of Biochemistry, 47, 45–148.Google Scholar
- 29.Wu, G., & Yan, S. (2002). Molecular Biology Today, 3, 55–69.Google Scholar
- 32.Yan, S., & Wu, G. (2010). Journal of Guangxi Academy of Sciences, 17, 145–150.Google Scholar
- 33.Wu, G., & Yan, S. (2008). Lecture notes on computational mutation. New York: Nova.Google Scholar
- 34.Feller, W. (1968). An introduction to probability theory and its applications, Vol. I (3rd ed.). New York: Wiley.Google Scholar
- 35.Hagan, M. T., Demuth, H. B., & Beale, M. H. (1996). Neural network design. Boston: PWS Publishing Company.Google Scholar
- 36.Demuth, H., & Beale, M. (2001). Neural network toolbox for use with MatLab. User’s guide. Version 4.Google Scholar
- 37.MathWorks Inc. (2001). MatLab—The Language of Technical Computing (version 18.104.22.1680, release 12.1). 1984–2001.Google Scholar
- 40.Sokal, R. R., & Rohlf, F. J. (1995). Biometry: the principles and practices of statistics in biological research (3rd ed., pp. 203–218). New York: W. H. Freeman.Google Scholar