Gene ontology analysis, expression analysis and SNP functional annotation

Please see the description below:

• Gene ontology analysis:
  The "enrichGO"  function in R package "clusterProfiler" (10.1089/omi.2011.0118) was used to conduct gene enrichment analysis. The 24 nearest genes of the top-associated SNPs at the 24 face-associated genetic loci were considered as the inputs. All ontology terms with nominal significance (p<0.01) after FDR correction were selected and categorized in four categories, 'morphogenesis', 'differentiation', 'development' and 'regulation' based on the process they involve.

  
  

• Expression analysis:
  RNA-seq data from 50 cell types were obtained from the study of Prescott et al., (10.1016/j.cell.2015.08.036, CNCC), GTEx and ENCODE (other 49 cell types). Normalized counts data were used for all cell types by applying VST approach. The comparative analyses included 1). the differences in normalized VST values between 24 face-associated genes and all genes over the genome in all cell types. 2). the differences in normalized VST values for the 24 face-associated genes between CNCC and other 49 cell types. Wilcoxon rank-sum test was used for statistical testing.

  
  

• SNP functional annotation:
  SNPs at the 24 face-associated genetic loci were annotated using the WashU Epigenome Browser (epigenomegateway.wustl.edu/browser/) based on epigenetic datasets including 6 tracks on Chip-seq (H3K27ac, H3K4me1 H3K4me3, p300, TFAP2A, and ATACseq) from Prescott et al. and conserved regions (PhyloP) from UCSC. Then the patterns of regulation within each associated region were investigated in details, where we calculated the log scaled quantile rank and variation of Chip-seq data for all LD SNP (r2 > 0.25) and base-pair in vicinity (within 20kb) to top-associated SNPs in 3 replicated measurements from 6 types of Chip-seq data. The top 5% significant results from any track were visualized.
  

Category