Understanding TCGA somatic mutation files

Question

[j.lunger18]

Hi all.

I'm currently looking at mutation data from TCGA. I have downloaded .json files from the GDC portal and parsed the files to download specific subsets of files.

I'm rather confused because the primary tumor somatic mutation .vcf files from TCGA are in a different format than I'm used to. The headers are as follows:

#CHROM  POS ID  REF ALT QUAL    FILTER  INFO    FORMAT  NORMAL  TUMOR

However, since I have indicated that I only want files from "primary tumor", what is this "NORMAL" column for?

I ask partly to understand, but partly because I want to eventually merge all the samples together downstream, and I'm not able to do so when all of the sample names look to be "NORMAL" and "TUMOR"...

Answer 1

[Kevin Blighe]

The normal column indicates the presence (or not) of each variant in the matched normal sample for the patient whose VCF you are viewing.

Kevin

[j.lunger18]

Will variants be listed that are only in NORMAL? Or does this entirely depend on the tool that was used for variant calling? I'm interested in looking at sites only, and I want the variants to only be those found in tumors.

[Kevin Blighe]

Will variants be listed that are only in NORMAL?

I do not believe this to be true; however, you can easily check, as such sites will have 0/0 for TUMOR.