Fig. 4: Motif implantation towards minimal enhancer design.

a, Prediction score distribution of the sequences for the γ-KC class (n = 2,000 sequences) after each motif implantation at best ___location (blue), random ___location (orange) and after 15 mutations (Nuc-15). The box plots show the median (centre line), interquartile range (box limits) and 5th and 95th percentile range (whiskers). b, Distribution of Mef2 locations relative to Ey (n = 2,000). c, Distribution of Onecut locations relative to Ey (n = 2,000). d, Prediction scores for motif implanted sequence (ME-1) after each motif implanting and generation of repressor sites. e, Nucleotide contribution scores of ME-1 in its initial form (first track) and after Ey, Mef2 and Onecut implantations (second track). Dashed lines show the position of the motifs. Delta prediction score for in silico saturation mutagenesis (third track). Black circles, selected mutations to generate repressor sites. Nucleotide contribution scores after generation of repressor sites (fourth track). Dashed lines show the position of the mutations. f,g, In vivo enhancer activity of the cloned 500 bp sequence with Ey, Mef2 and Onecut implantations (f) and after generation of repressor sites (g). h, Zoom into the selected 49 bp part of the 500 bp sequence from e. The size of the motifs, the spaces between motifs and the flankings are shown at the bottom. i, In vivo enhancer activity of the cloned 49 bp sequence with Ey, Mef2 and Onecut implantations. In f,g,i, the expected ___location of γ-KC is shown with dashed circles. Scale bars, 100 µm. Ey (E), Mef2 (M), Onecut (O) and Sr (S).