How exactly does the nonpeaks file differ among the different folds? ie from the script it seems like train/test/valid info is not used for creating gc matched negatives. After generating the negatives file I am noticing all the chromosomes are still present in all the folds so do I have to create the nonpeaks file per fold or can I just use the same nonpeaks file for every fold?
How exactly does the nonpeaks file differ among the different folds? ie from the script it seems like train/test/valid info is not used for creating gc matched negatives. After generating the negatives file I am noticing all the chromosomes are still present in all the folds so do I have to create the nonpeaks file per fold or can I just use the same nonpeaks file for every fold?