This paper focuses on the comparison of two different approaches to the analysis of Single Nucleotide Polymorphism (SNP) profiles data regarding Crohn's Disease; the first one is based on a single SNP analysis, conducted by means of classical statistical tools, to assess the correlation existing between SNP's profile and phenotype; the second one makes use of classifiers based on Regularized Logistic Regression. The findings of the study show that the machine learning techniques adopted are able to provide statistically significant prediction accuracy of the phenotypic status of the subjects analyzed by SNP data. Moreover, they are poorly influenced by the noise embedded in the data and are suitable for genome-wide analysis. © 2008 Springer-Verlag Berlin Heidelberg.
Prediction of Crohn's Disease by profiles of single nucleotide polymorphisms / Colella, R.; D'Addabbo, A.; Latiano, A.; Palmieri, O.; Annese, V.; Ancona, N.. - 5179:3(2008), pp. 564-571. ( 12th International Conference on Knowledge-Based Intelligent Information and Engineering Systems, KES 2008 Zagreb, hrv 2008) [10.1007/978-3-540-85567-5_70].
Prediction of Crohn's Disease by profiles of single nucleotide polymorphisms
Annese V.;
2008-01-01
Abstract
This paper focuses on the comparison of two different approaches to the analysis of Single Nucleotide Polymorphism (SNP) profiles data regarding Crohn's Disease; the first one is based on a single SNP analysis, conducted by means of classical statistical tools, to assess the correlation existing between SNP's profile and phenotype; the second one makes use of classifiers based on Regularized Logistic Regression. The findings of the study show that the machine learning techniques adopted are able to provide statistically significant prediction accuracy of the phenotypic status of the subjects analyzed by SNP data. Moreover, they are poorly influenced by the noise embedded in the data and are suitable for genome-wide analysis. © 2008 Springer-Verlag Berlin Heidelberg.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
