Hello!
I've been looking to adapt dndscv for use with mouse (mm10). I appreciate the availability of the genome annotations, but had wanted to apply a covariate matrix, given my testing with human data suggests it really helps with the model.
Namely, I wanted to leverage ENCODE mouse chip datasets to do what you did in the 2017 paper, and wanted to ask how this was done. Is it like average sequence coverage for a given mark (e.g. H3K27me3) in a gene body across different samples? Then the eventual matrix is gene x mark, and you run a PCA on this to get the covariate matrix?
Thanks!
Daniel
Hello!
I've been looking to adapt dndscv for use with mouse (mm10). I appreciate the availability of the genome annotations, but had wanted to apply a covariate matrix, given my testing with human data suggests it really helps with the model.
Namely, I wanted to leverage ENCODE mouse chip datasets to do what you did in the 2017 paper, and wanted to ask how this was done. Is it like average sequence coverage for a given mark (e.g. H3K27me3) in a gene body across different samples? Then the eventual matrix is gene x mark, and you run a PCA on this to get the covariate matrix?
Thanks!
Daniel