Closed mdozmorov closed 9 years ago
Right now, we have 'bulk' FDR adjustment, by adjusting each column of the matrix_PVAL.txt.
We need to confine FDR correction to each cell type separately. E.g.,
E001-H3K4me1 0.005 E001-H3K4me3 -0.3 E033-DNase 0.004 E033-H3K4me1 0.2
should be treated considering E001, E033 assignment. That is, 0.005, -0.3 should be FDR corrected separately from 0.004, 0.2.
Implement this with help of 'gf_descriptions.txt' file, where we have cell type assignments. It is currently implemented in https://github.com/mdozmorov/R.genomerunner/blob/master/utils2.R#L423
Obviously, some GFs, like 'TFBS_clustered', will have one "cell type" - 'TfbsV3'. So, they will be corrected within this cell type.
Right now, we have 'bulk' FDR adjustment, by adjusting each column of the matrix_PVAL.txt.
We need to confine FDR correction to each cell type separately. E.g.,
E001-H3K4me1 0.005 E001-H3K4me3 -0.3 E033-DNase 0.004 E033-H3K4me1 0.2
should be treated considering E001, E033 assignment. That is, 0.005, -0.3 should be FDR corrected separately from 0.004, 0.2.
Implement this with help of 'gf_descriptions.txt' file, where we have cell type assignments. It is currently implemented in https://github.com/mdozmorov/R.genomerunner/blob/master/utils2.R#L423
Obviously, some GFs, like 'TFBS_clustered', will have one "cell type" - 'TfbsV3'. So, they will be corrected within this cell type.