Figure 2.

The Relationship between Promoter CpG Density (o/e CpG Ratio) and Loss-of-Function Intolerance Conditional on Downstream Gene Expression Level and Tissue/Developmental Specificity (τ)

(A) The distribution of LOEUF, stratified by promoter CpG density, in each quartile of downstream gene expression level, computed using the GTEx dataset (Material and Methods).

(B) The distribution of LOEUF, stratified by promoter CpG density, in each quartile of downstream tissue specificity. For each gene, tissue specificity is quantified by τ, and is computed using the GTEX dataset (Material and Methods).

For both (A) and (B) quartiles are labeled from 1-4, with 1 being the quartile with the lowest and 4 the quartile with the highest expression/tissue specificity, respectively.

(C) The percentage of LOEUF variance (adjusted r²) that is explained by downstream gene expression level, τ, the interaction between the two, and promoter CpG density.

(D) The distribution of LOEUF, stratified by promoter CpG density, for 559 genes whose mouse homologs are differentially expressed at specific time points during embryogenesis (Material and Methods). The stratification was done based on the CpG density quartiles calculated for all 4,743 genes, as in (A) and (B).