Cancer Classification Using Loss of Heterozygosity Data Derived from Single-Nucleotide Polymorphism Genotyping Arrays

Single-Nucleotide Polymorphism (SNP) array is a recently introduced technology that genotypes more than 10,000 human SNPs on a single array. It has been shown that genome-wide Loss of Heterozygosity (LOH) calls can be derived by analyzing the genotypes calls measured by SNP arrays using paired tumor and normal tissue samples. The goal of this study is to evaluate the possibility of cancer classification using LOH calls. As a proof of concept, we applied 16 different combinations of classification algorithms and feature selection algorithms to a public data set that contains LOH calls of 10,043 SNP loci obtained from 10 breast cancer patients and 5 small cell lung cancer (SCLC) patients. Performance was measured in terms of the leave-one-out cross-validation (LOOCV) classification accuracy. Experimental results suggest that LOH calls derived from SNP arrays can be an excellent indicator of cancer type