I have downloaded all the VCF files for one cancer type from the TCGA. Upon annotation using Annovar, I notice that there are a lot more variants than I expected. So I have a few questions:
- How should I process these downstream? I've already reduced the list based on those that passed.
- I particularly wanted the mutation profiles of individuals before MutSig. Is using the VCF files the best way to do this, or should I just get the MAF file? My assumption was that the MAF file was MutSig processed.
Thank you.
1 answer
TCGA VCFs are raw mutation lists, straight out of the automated variant calling pipelines. TCGA MAFs are also mutation lists, but have (usually) undergone expert curation to remove false-positives, or recover known calls missed by the automated pipelines. This post should give you a good intro to TCGA VCFs and MAFs - Working with MAF files (Mutation Annotation Format) from the TCGA (The Cancer Genome Atlas). The MAF is usually the best place to start for good science, unless you are confident that the VCFs contain data you need, that was removed in the corresponding MAF. Take a look at this post to find and re-annotate the MAF for your cancer type - Annotating TCGA MAFs with the latest Ensembl/Gencode transcripts
Log in to answer this question.
Hi gaiusjaugustus, I found many vcf files are controlled. Can you tell me how to get those data? Many thanks.
NCI is transitioning to use GDC for access to the same protected datasets, with slightly less red tape. See https://gdc.nci.nih.gov/access-data
Thanks Cyriac Kandoth. As from Australia, I am not in any TCGA (or NIH) official collaborators, how can I get access to those controlled VCF or MAF data, for example https://gdc-portal.nci.nih.gov/files/01776835-76e8-4938-83da-710a03b49221
MAFs are the input to MutSig. They are not MutSig processed.