The files in this directory contain chromHMM chromatin state calls for multiple tissues during mouse embryonic development, as described here: http://www.biorxiv.org/content/early/2017/07/21/166652.

All files are in bed format. There is one file for each tissue at each developmental stage.
The columns are 1) chromosome, 2) regions start, 3)  region end, 4) chromatin state number 5) Chromatin state label. 
There are 15 chromating states, labeled as follows:
State	Label	Description
1	Pr-A	Promoter, Active
2	Pr-W	Promoter, Weak/Inactive
3	Pr-B	Promoter, Bivalent
4	Pr-F	Promoter, Flanking
5	En-Sd	Enhancer, Strong TSS-distal
6	En-Sp	Enhancer, Strong TSS-proximal
7	En-W	Enhancer, Weak TSS-distal
8	En-Pd	Enhancer, Poised TSS-distal
9	En-Pp	Enhancer, Poised TSS-proximal
10	Tr-S	Transcription, Strong
11	Tr-P	Transcription, Permissive
12	Tr-I	Transcription, Initiation
13	Hc-P	Heterochromatin, Polycomb-associated
14	Hc-H	Heterochromatin, H3K9me3-associated
15	NS	No Chromatin Signal
NRS	NRS	No Reproducible State (ony relevant to "replicated" call set)


There are four subdirectories:

rep1\
This subdirectory contains chromatin state calls made using ChIP-seq datasets from biological replicate 1 (two biological replicates were performed for each ChIP-seq experiment).

rep2\
This subdirectory contains chromatin state calls made using ChIP-seq datasets from biological replicate 2.

replicated\
This subdirectory contains the intersection of the replicate 1 and replicate 2 chromatin state calls. 
Here, we require that a region is called in the same state in both replicates. If a region is not called in the same state in both replicates, it is labeled "NRS" for No Reproducible Signal. Note this is differenct than state 15, which is "No Signal." 

pooled\
This subdirectory contains chromatin state calls on ChIP-seq data pooled from both replicates.
We expect this will be the desired set of files for most users.


Each subdirectory has two additional subdirectorites:

archive\
This contains an older version of the bed files in which there are only state number, no labels. Aside from the labels, the files are identical to the working versions in the main directory.

tracks\
This contains bigBed versions of the files for visualization on the UCSC Genome Browser.


There is one additional subdirectory that does not conform to the templates decribed above. This is 6_mark_models\.
This subdirectory contains chromHMM segmentation of the e10.5 stage where we did not have all 8 marks. 
Note that the 11-state and 16-states models used for this segmentation are distinct from the 15-state (8-mark) model described above for the other stages. 
Please see manuscript and Extended Data Figure 7 for additional information.