Chaos Embedded Particle Swarm Optimization for Tag Single Nucleotide Polymorphism Selection

Single Nucleotide Polymorphisms (SNPs) are the most common variants in the human genome. Disease analysis costs can be reduced by selecting meaningful SNPs, i.e., tagging the SNP selection. We propose a method, called chaos particle swarm optimization (CPSO), to select tag SNPs, and use linkage disequilibrium (LD) and the K-nearest neighbor (K-NN) method to respectively reduce and evaluate the tag SNPs. To measure the quality of the correction rate and the tag SNPs number, the Hap Map database was used to test CPSO´s ability and to compare the proposed method with other methods. The results indicate that the proposed method is effectively to enhance the tag SNP prediction in terms of the result achieves a good accuracy when compared to methods from the literature.