Fwd: homozygous mutations or not

[JianJiong Gao]

Forgot the cc the group.

---------- Forwarded message ----------
From: JianJiong Gao <jg...@cbio.mskcc.org>
Date: Thu, Mar 27, 2014 at 10:18 PM
Subject: Re: homozygous mutations or not
To: Solip Park <imag...@gmail.com>


Hi Solip,

I think this type of analysis is tricky. I had done some analysis like
that but I don't really have a lot of insight on this. Tumor purity,
clonality and ploidy could all affect allele frequency. For
multi-clonal and aneuploid samples, it is complicated. For the first
analysis, I would only consider clonal and deploid samples.

For clonal pure tumor samples, the allele freq should peak at 0.5 with
a long tail  some mutations near 1.0 -- so it's easy to call, e.g.
taking allele freq greater than 0.9 as homozygous. For impure samples
but still with one clone, maybe you could compared the allele freq of
a mutation to the mean allele freq of all mutations in a sample, and
take a threshold on the ratio to call homozygous.

Currently, in the cBioPortal, we don't call purity, clonality or
ploidy. You may need to look into each individual sample. For example
in this case (http://www.cbioportal.org/public-portal/case.do?cancer_study_id=ucec_tcga_pub&case_id=TCGA-AP-A0LF),
it seems to be an impure tumor and I would call TP53: R273S as a
homozygous mutation.

I wish I could be more helpful. Please let me know if you have any
thoughts. Maybe we could use your ideas to import analysis on the
cBioPortal too.

Best,
-JJ



On Tue, Mar 25, 2014 at 12:57 PM, Solip Park <imag...@gmail.com> wrote:
> Hello JianJiong :)
>
> How are you?
>
> As always, I have one question about cBioportal. Now I have two data sets
> from your website (copy number variants;GISTIC and mutation data; MAF).
> This time, I want to know that my query mutations are homozygous mutations
> or not. Is there anyway to decide the type of mutations (homozygous or
> heterozygous) from your website?
>
> Because your website provides "allele freq", so I guess there is a way to
> detect this.
> Of course, I know the "absolute" will be a solution, but I want to know
> there is another way to do this.
>
> Always many thanks :)
>
> Solip
>