README

We offer four types of data files for the H3K4me1/me3 study: 
a) compressed (gzipped) WIG (http://genome.ucsc.edu/google/goldenPath/help/wiggle.html) format files that 
   can be loaded directly into the UCSC genome browser to show coverage profiles.
b) compact, tabular, tab-delimited 'peaks' files that support high throughput calculations and can readily be 
   transformed into UCSC-compatible BED-format files (http://genome.ucsc.edu/goldenPath/help/hgTracksHelp.html#BED). 
c) SEQ and PRB files generated by the base calling section of the Illumina data processing pipeline. 

For example, the HeLa 'H3K4me1 untreated' folder contains: 
H3K4me1_unstim_hg18_xset200_dupsN_ht7.sub.peaks  23-Jul-2008 18:06  7.9M  
H3K4me1_unstim_hg18_xset200_dupsN_ht7.sub.wig.gz 23-Jul-2008 18:06   35M  
H3K4me1_unstimulated_1_prb.txt.gz                24-Jul-2008 11:14  1.1G  
H3K4me1_unstimulated_1_seq.txt.gz                24-Jul-2008 11:11  123M
H3K4me1_unstimulated_2_prb.txt.gz                01-Apr-2008 17:41  1G  
H3K4me1_unstimulated_2_seq.txt.gz                24-Jul-2008 11:11  125M  
H3K4me1_unstimulated_3_prb.txt.gz                24-Jul-2008 11:13  1.4G  
H3K4me1_unstimulated_3_seq.txt.gz                24-Jul-2008 11:10  173M  
H3K4me1_unstimulated_4_prb.txt.gz                24-Jul-2008 11:14  1.4G  
H3K4me1_unstimulated_4_seq.txt.gz                24-Jul-2008 11:11  175M

For WIG and 'peaks' files, filenames indicate the assembly against which reads were aligned (hg18 or mm8) 
and the XSET extension length (200 bp) used by FindPeaks 2.0 to generate the coverage profile. Both types of 
files contain information only for enriched regions that have a maximum XSET overlap value that is at least 
as significant as the FDR ~0.01 threshold that we estimated independently for each Eland read set 
(for the above dataset, the threshold height was 7). While the thresholded WIG files display no information 
below the FDR threshold, they are compact enough that the whole genome dataset can be loaded as one file, 
which facilitates browsing genomic locations on different chromosomes. 

The '*.peaks' files use one-based start coordinates, rather than being zero-based, half-open 
(http://genome.ucsc.edu/google/FAQ/FAQtracks#tracks1). The files contain the following six fields:
<chromosome> <genomic start> <genomic end> <position of peak maximum relative to genomic start> <peak height> <number of reads in the peak>

For SEQ and PRB files, the number in the filename is an ID for a flow cell lane. 
In the above example, H3K4me1_unstimulated, there were 4 lanes of data. 

All data are single-end reads that were generated on Illumina 1G sequencers.