scales.R
calculate scaling factors for each sample, gives scales.csv file used for bedgraphs2.sh 

bedgraphs2.sh
produce bedgraph files from bamfiles using genomeCoverageBed.  These are sorted and filtered using sed
to get rid of anything not mapped to 1-22XY.

The bedgraph files have a value for each interval defined by read ends.  This is the equivalent of having a 
value for every nucleotide. The counts are per million reads and also divided by the number of samples in the group, to 
the sum we get at the end is a weighted mean.

union.sh
multicolumn bedfraph files for control and AD produced using unionbedg.
inputs are  bedgraph/${sample}.4bgs

sumum*.sh
bw files generated by summing union bedfiles (perl script 'summit'),
then bedGraphToBigWig

bw files in trackhub on webserver
/home/www/ucsctrackhub/AD_H3K27ac_experimental/hg19/trackDb.txt
...
track ADBedtools
bigDataUrl ADunionS.bw
...
track controlBedtools
bigDataUrl conunionS.bw
...