Read counts and target regions identified by ChIP-Seq.

1 Uniquely mapped targets were identified by PeakSeq and further filtered using criteria more stringent than the default parameters. See Materials and Methods.2 Genic regions were identified for the total number of targets determined by PeakSeq. Genic regions are defined as a window encompassing 2.5 kb up- and downstream of the 5′ and 3′ ends, respectively, using protein-coding genes from Ensembl build 52 based on hg18.3 The high-confidence union list contains 49,555 targets and was formed by creating a union list of all SWI/SNF regions from those identified by PeakSeq and trimming this list to those SWI/SNF regions that co-occur with each other, Pol II regions, 5′ ends of protein-coding genes, CTCF sites or putative enhancers. For further details, see Materials and Methods.