Resequencing of 12 CEPH individuals based on captured genomic DNA

We designed an Agilent 244k SureSelect microarray for targeted resequencing on the 51,686 ChIP-seq peaks identified in the 12 samples. We used fimo (http://meme.sdsc.edu/meme/) to scan for instances of the core 14 bp of the canonical motif [19] with a p-value of 10-2. We adjusted the target locations to center on matches to the nearest CTCF motif if the motif was within 50 bp, and added flanking targets to capture additional nearby motifs. 5 potential probes were tiled at 15 bp spacing to the 120 bp surrounding each target. We adjusted probe binding energy similarly to Ng et al. 2009 [54], adjusting the spacing of probes by up to 5 bp and adjusting the lengths to between 40-60 bp to reach a predicted Tm between 60-72 °C. We used the Duke Uniqueness 20 bp track (UCSC genome browser) to filter out 5,828 probes with potential for cross-hybridization. We further excluded 145 probes in satellite repeats (RepeatMasker, UCSC genome browser) or with high blast scores to multiple genomic locations. The final design had 242,380 probes targeting 46,652 CTCF sites.