Impact of multi-reads on sequencing depth.

In the first column, “(C)” and “(I)” refer to ChIP and input samples, respectively. Percentages in the third to fifth columns are calculated with respect to the total number of reads (the second column). The actual numbers of reads are provided in Table S1. “% Rescued” in the last column is obtained as the number of multi-reads divided by the number of uni-reads and it indicates the gain in sequencing depth due to multi-reads.