Topography analysis of signatures

Reference information of replicative strands and replication-timing regions were obtained from Repli-seq data of the ENCODE project (https://www.encodeproject.org/)⁸⁵. The transcriptional strand coordinates were inferred from the known footprints and transcriptional direction of protein coding genes. First, for a given mutational signature, one could calculate the ‘expected’ ratio of mutations between transcribed and non-transcribed strand, or between lagging and leading strands, according to the distribution of trinucleotide sequence context in these regions. Second, the ‘observed’ ratio of mutations between different strands can be identified through mapping mutations to the genomic coordinates of all gene footprints (for transcription) or leading/lagging regions (for replication). Third, all mutations were orientated towards pyrimidines as the mutated base (as this has become the convention in the field). This helped denote which strand the mutation was on. Fourth, the level of asymmetry between different strands was measured by calculating the odds ratio of mutations occurring on one strand (e.g., transcribed or leading strand) vs. on the other strand (e.g., non-transcribed or lagging strand). IntersectBed⁸⁶ was used to identify mutations overlapping certain genomic features.