The genetic algorithm for breast tumor diagnosis-The case of DNA viruses

The purpose of this study is to determine the bioinformatics between breast tumors and DNA viruses. To extract significant factors for breast tumors, two methods - analysis of variance (ANOVA) and information measurement - are used; then, the results show that human papillomavirus (HPV) is not significant factor. Further, genetic algorithm (GA)-based data mining was submitted to obtain the rule between other DNA viruses and breast cancer/fibroadenoma. The GA rule shows that the effect of breast cancer includes {herpes simplex virus (HSV)-1 (-), human herpesvirus (HHV)-8 (-,+)}, {HSV-1(-)}, or {HHV-8(-)}, and that the effect of fibroadenoma includes {HSV-1(+), HHV-8(+)}, {HSV-1(+)}, or {HHV-8(+)}. Furthermore, the Mantel-Haenszel test shows that GA rules make a distinction between breast cancer and fibroadenoma. Finally, regarding diagnosis cost and accuracy, this study suggests that considering HSV-1 alone or with HHV-8 is optimal, but considering only HHV-8 yields less accurate results.