Estimate contamination in Next Generation Sequencing data

[desmo]

Hi everyone,
I'm a bioinformatician and I'm working on cancer genomes.
I would like to know how I can estimate the contamination rate both of the tumor and normal sample of the patient that we will have to sequencing.
I know that ContEst could work for my problem but we didn't have any genotype data.
So there are some methods that could be applied to sequencing data to solve this problem?
Thanks in advance

[JackieBadger]

I think the best way is to use relative depths, and fit your data to expected models. In your case it is quite simple with diploid genotypes. True alleles should be at much greater depth than artifacts/contamination.

We see contamination between multigene amplicons, and use a model approach. Still when contamination is significant the genotype can not be estimated.

A recent paper talks a little about the issue: http://genomebiology.com/content/13/5/R34