Integrative approaches in nsCL/P

1) cNCC_UCSC_ChromHMM.bed, and hNCC_UCSC_ChromHMM.bed.bed These files contain the chromatin segmentation data that was generated based on chromatin immunoprecipitation followed by sequrencing (CHiP-seq) data of different histone modifications captured in early human neural crest cells (hNCC, Rada-Iglesias et al. 2012 (GSE28874)) and cranial neural crest cells (cNCC, Prescott et al. 2015 (GSE70751)). While the histone marks H3K27ac, H3K4me1, H3K4me3 and H3K27me3 were directly measured in hNCC, additional histone marks (H3K9me3 and H3K36me3) were imputed using ChromImpute v1.0.1 (Ernst and Kellis, 2015) based on 127 cell types from Roadmap Epigenome Project (Roadmap Epigenomics Consortium et al. 2015) and the available chromatin marks in hNCC, cNCC and craniofacial tissue before chromatin segmentation. For the chromatin segmentation itself, we used the core 18 + H3K27ac model as implemented in ChromHMM (Ernst and Kellis, 2012). The files are provided in a genome browser compatible format. 2) CT_active_NCC_inactive.bed, and CT_act_NCC_inact_UCSC.bed These files contain specific active sites in NCC or CT. In the chromatin segmentation datasets, we filtered for active chromatin sites (Transcription starting sites (TSS), Enhancer or transcribed sites) in NCC that are repressed/quiet (Quiescent, Bivalent_TSS_poised_Enhancers, RepressedPolyComb sites, Heterochromatin) in CT and vice versa. For robust observations, we only trust in a chromatin state if it is present in both NCC samples (hNCC, cNCC), or in five of the six CT (Carnegie stage (CS) 13, CS14, CS15, CS17, CS20, 1 weeks post-conceptum) samples. To account for biases in length associated with batch-effects, active sites were only retained if they had a distance of ≥500 base pairs to any chromatin segment of opposite activity status in the other cell system/tissue. In the following, we combined the specific active chromatin sites with associations of our meta-analysis in cleft (MAiC) and topologically associating domain (TAD) data to filter for TADs with high density of strong associated genetic variants (PMAiC ≤ 5×10-05) in specific active chromatin sites at new and known nsCL/P GWAS risk loci.