The directories contain imputed ChIP-seq data from ChromImpute.

For ChIP-seq the data are in bigwig (.bw) files and correspond to -log10 p-val tracks at a 25 base pair resolution.
 
Imputed data for a mark in an epigenome is only available if there is at least one other mark with ChIP-seq data in that epigenome.
Here are the total number of epigenomes with imputed data for each histone modification out of the 1704 that were considered:
H3K27ac - 1088
H3K27me3 - 1703
H3K36me3 - 1703
H3K4me1 - 1688
H3K4me3 - 1688
H3K9me3 - 1700

In addition due to QC concerns after imputation the following epigenomes ChIP-seq were dropped from the integrative analysis
IHECRE00000833, IHECRE00003135, IHECRE00003202, IHECRE00003229, IHECRE00003326, IHECRE00003355
Note IHECRE00003135, IHECRE00003202, IHECRE00003229, IHECRE00003326, IHECRE00003355 did not have imputed data for H3K27Kac since that was the only mark in those epigenomes. 

Only chr1-22 and chrX were imputed.

Within each sub-directory there is a .md5 with the checksum for all the .bw files in the subdirectory.

More information about ChromImpute can be found here:
Ernst J, Kellis M. Large-scale imputation of epigenomic datasets for systematic annotation of diverse human tissues. Nature Biotechnology, 33:364-376, 2015.

Contact Jason Ernst (jason.ernst@ucla.edu) with any questions about the imputed data.