How is MutSigCV's example LUSC data generated?
I'm having difficulty working with MutSigCV's example LUSC data found at the bottom of this page: https://www.broadinstitute.org/cancer/cga/mutsig_run
I noticed the following:
(1) There are many intronic mutations in the example data that are not "covered" (i.e. did not have sufficient read-depth for mutation calls) according to the coverage WIG files of the corresponding TCGA samples (output by MuTect) obtained via Broad's Firehose interface.
(2) There are intronic mutations in the example data that are "covered" in the Firehose data, but not called as mutations (i.e. they weren't in the MAF files provided by TCGA).
Given this, I have the following questions:
(1) Is the example LUSC data generated from the same read datasets (BAM files) as the ones provided by TCGA? If so, any thoughts on what might have caused the above discrepancies?
(2) Given that the coverage WIG files in Firehose are not consistent with the mutation calls in the example LUSC data, is there any place where I can find the right WIG files?
Any help would be greatly appreciated!