Extended Data Fig. 1: Properties of double peak metaprofile.

a) The frequency of distance between the two maxima in a double peak flanking active STARR-seq peaks is plotted. b) The symmetricity of the double peak pattern is plotted. The ratio of the distance between the two peaks to the ratio between one of the maxima and the minima is plotted. While there is large amount of variability in the distance between the two peaks (mostly between 300-1100 bp), the trough in the double peak tends to occur in the center of the two peaks. (c) The metaprofile around active STARR-seq peaks is plotted for different epigenetic marks. Histone marks that are enriched near STARR-seq peaks display the characteristic double peak pattern shown in c) due to the depletion of histone proteins at active regulatory regions. In addition, DHS displays a single peak at the center of these regulatory regions as shown in c). d) On the other hand, no such double peak pattern is observed on depleted histone marks at STARR-seq peaks. e) The matched filter score is calculated using the convolution of the flipped metaprofile and the epigenetic signal using a sliding window of variable length. The significant peaks in the final matched filter score are used to identify active regulatory regions.