Thanks. I am looking to test how PWM generated from different sources (PBM, HT-SELEX, ChIP-chip, Promoters, Footprinting, e.t.c.) can be able to discriminate sequences with known binding sites from the background. So I want to expand the collection of ChIP-seq data that I have. These are the data from ENCODE that I am currently using.
You should try ChIPBase (details are in this post), a comprehensive database for interpreting ChIP-seq data from multiple source, including ENCODE, GEO database. They also provided identified DNA-binding motifs and their regions of TFs.