Table 1. Comparison of CREST-seq to published CRISPR/Cas9 screens of non-codin gregulatory sequences.

CREST-seq is compared to the published screening of non-coding regulatory elements. The following aspects are compared: the size of screen region; the total number of oligos required to construct the library; the average number of oligos per kilobase in each screen; and the estimated coverage of the target region. To estimate the coverage of target region, we assume that the PAM motifs are equally distributed across the genome and each gRNA creates a mean indel size 9.5bp ± 13.7bp. To compute the coverage of CRISPRi screen using dCas9-KRAB, we assume that the average size of H3K9me3 peaks introduced by dCas9-KRAB is about 850bp.

Reference	Target region	Total oligo #	Oligo density (per KB)	Coverage	Distinguish Trans or Cis?
Canver et al, 2015 10	4.2kb, 3 DHS and 1 exon	582	137	~1x	No
Korkmaz et al, 2016 11	685 p53 ChIP-seq peaks	1116	NA	1.3-1.6 oligos per ChIP-seq peak	No
	73 eRNA expressing ERa ChIP-seq peaks	97	NA
	2kb, deCDKN1A locus	197	98.5	<93.6%
Rajagopal et al, 2016 12	40kb Tdgf1 locus	3908	98	<93.1%	No
	Rpp25, Nanog, and Zfp42 loci	3908	NA	NA
Diao et al, 2016, 13	37.6kb, 174 putative enhancers in 1Mbp POU5F1 locus	1964	52	<49.4%	No
Sanjana et al, 2016 14	200kb NF1 locus	6682	33.4	<31.7%	No
	200kb NF2 locus	6934	34.6	<32.9%
	200kb CUL3 locus	4699	23.5	<22.3%
Fulco et al, 2016 15	1.29Mbp, GATA1 and MYC loci	98,000	76	~64x	No
CREST-seq	2Mbp, POU5F1 locus	11,600	5.7	20x	Yes