Data used in this study.

This table describes the data used for this study. The main data includes five cell lines from both University of Washington (UW) and Duke University (Duke). Bias from this data has been illustrated in Figure 1 panels w1–w5 and d1–d5. The supporting data rows describes data from four more experiments that either followed the UW or Duke protocols. The bias of these data is illustrated in Figure S1 panels I-III and IV-V. The control data is built from two ChIP-Seq histone modification data sets. We illustrated results of our analysis based on these two data sets in Figure 1 w6–w7. In the protocol column, those data generated in UW and Duke are denoted as UW and Duke respectively, but those which are a slight modification of these protocols are denoted as UW-like and Duke-like. The file names column shows the name of the files that we obtained from the ENCODE repository at UCSC and used in this study.