MAF files MutSigCV runs

[KAYA Arife Neslihan (SINGA)]

Dear Madam/Sir,

I have tried to find driver genes in KIRC using MutSigCV (version 1.41) and downloaded MAF files from FireBrowse. First I tried using raw Mutation Annotation files (merged all MAFs for all 451 patients and removed one of the samples if the patient has more than two tumour sample barcode). MutSigCV identified 207 driver genes with q<0.1.  When I try using "KIRC-TP.final_analysis_set.maf" which you used in MutSigCV analysis, number of drivers is 86 and overlap between the previous run is only 40. Number of drivers in your analysis using "KIRC-TP.final_analysis_set.maf" is  149.  I also experienced similar issues for LUAD. Could you help me to understand these huge differences between analysis results?

Sincerely,

Neslihan

This e-mail and any attachments are only for the use of the intended recipient and may contain material that is confidential, privileged and/or protected by the Official Secrets Act. If you are not the intended recipient, please delete it or notify the sender immediately. Please do not copy or use it for any purpose or disclose the contents to any other person.

[David Heiman]

Hi Neslihan,

Unfortunately, it is not possible for an outside user to run MutSigCV in the same fashion we have on Firehose, as part of our process applies a blacklist that includes undocumented germline data that we are unable to release.

The KIRC-TP.final_analysis_set.maf file is actually the file generated by MutSigCV after applying the blacklist and other filters.

Our team does not develop MutSigCV, we used run the tool on their behalf. If you want to try and get in touch with the developers, you can try the CGA software help forum.

MutSigCV has been out of active development for a long time. If you are simply looking for MutSig analysis results on TCGA cohorts, I would suggest looking at firebrowse.org, where we have run the latest version of MutSig (2CV) on all the TCGA tumor types.

Regards,

David

--

David Heiman

Run Operations Engineer

TCGA Genome Data Analysis Center

The Broad Institute of MIT and Harvard

415 Main Street

Cambridge, MA 02142

[Weiwei Zhai]

Dear David, Thank you for the information and advice on MutSigCV results.

We browsed through the firehose(firebrowse) results, the number of identified

drivers are quite high (e.g. for lung adenocarcinoma, there are more than 1000

drivers). This is quite atypical.  We are wondering whether there is anything

wrong on ur end running the pipeline.

Since we don't have the engine you guys are using, it is hard for us to figure

out the causes of this. This is the major reason we wrote this email.

Would appreciate ur advice and help on this inquiry.

Thank you!

Weiwei

[Broad Institute GDAC]

Hi Weiwei.

The mutation data we use in Firebrowse is raw data from the TCGA. This means we do not control the callers used, nor have we done any data curation other than ensuring all annotations are Gencode 19 via the use of Oncotator. Artifacts and noise from uncurated data can have deleterious effects on mutation significance analysis.

MutSig CV has been deprecated as a tool. We strongly recommend you use the results of MutSig 2CV for SMGs, as it is much more robust (54 SMGs for LUAD).

Regards,

David

[Weiwei Zhai]

Dear David,  Thank you for the reply. We looked a bit deeper

and there seems to be a huge difference between mutsigcv and mutsig2CV.

for example, mutsigCV identifies 1498 drivers for LUAD.

We wrote to Michael Lawrence for the mutsig2CV. I am not sure

we can get a reply from him (perhaps too many requests of this sort).

It is a pity we cannot access this new tool. :(

Thank you for your kind reply again!

Weiwei