Association study with many types of variants

Question

[weiyuchung]

Hi all,

There are many developed methods for specific variant (e.g. SNPs, Indels or CNVs)　in association study. However, the disease may not just be attributed to one of the specific variant. Based on the NGS technique, we can identify many types of variants simultaneously. Is it reasonable to combined all the variants signals in one association study algorithm? Is there any comprehensive existing method to do so whether using raw reads information or variants calling?

Thanks