How to cluster single nucleotide polymorphisms that appear together ?

Question

[ceruleanivy]

I am running an analysis on R for ER,PR and HER2+ breast cancer patients, looking to investigate commonly repeated SNVs according to the molecular subtype. I might test intratumoral heterogenity as well by comparing front and core sample.

Given that my SNP list is inserted as columns with dummy variables (1, 0) I've tried the following so far

• Principal Component Analysis
• A priori (association rules)

What I don't know is if the above statistical methods are supported for that kind of analysis (papers ?). I would also appreciate alternative ideas, other than the two, on how to approach the subject.

[poisonAlien]

Check out sciClone.