Sinusoidal Map Based Particle Swarm Optimization Detect the SNP Barcode in Breast Cancer to Disease Susceptibility

Abstract

Single nucleotide polymorphisms (SNPs) are the most common type of DNA sequence variation in the human genome and are widely used to investigate the association analysis of diseases. SNP barcode is a combination of SNPs with genotypes (AA, Aa, and aa for an SNP) to find the difference between case data set and control data set for analyzing the disease association amongst SNPs. Currently, the computational time of statistical method becomes the weak to analyze the big data to find the significant SNP barcode. Here, we applied a sinusoidal particle swarm optimization (SPSO) algorithm facilitate the statistical methods to analyze the associated SNPs. We systematically evaluated the synergistic effect of 26 SNPs from eight epigenetic modifier-related genes in breast cancer. The 2- to 5-order SNP barcodes were found to determine the risk effects in breast cancer. We found that five of eight genes (BAT8, DNMT3A, EHMT1, DNMT3A, and BAT8) were statistically significant to breast cancer and play the important role in the SNP barcode. In addition, we compared the search ability between PSO and SPSO in the 2- to 5-order SNP barcodes. The results indicated that SPSO can find the better SNP barcode than PSO. In conclusion, SPSO is a precise algorithm for finding a significant model of SNP barcode.