An Improved Collapsing Method Of Detecting The Association Of Rare Variants And Common Diseases

Recently, although genome-wide studies had achieved great success in detecting common variants of common diseases by tagSNPs, it did not perform well in detecting rare variants. According to recent studies, common diseases were caused by rare vari-ants. The technology of testing DNA sequence directly can only detect limited diseases. With the developing of DNA sequence detecting technology, it had developed very well into the whole DNA sequence. Many methods of testing the association of rare vari-ants and common diseases had been proposed in the past, but it did not perform as well as we expected. However, the collapsing method that Li&Leal (2008) had proposed performed quite well. Its power was far greater than any other methods in the past. Also, its type I error was well controlled. However, its test was based on Pearson’s test which contained limited sample information. Based on CAST method proposed by Thilly, who made full use of the sample information, could not achieved the same power as Li&Leal. This paper combined the two methods together and taken advan-tage of them, proposed an improved CAST method. By simulation, it turned out that the proposed method performed better than the existing methods such as Single-maker test, Multi-maker test, Collapsing&Pearson method and Bayesian analysis test.