Journal of Genomes and Exomes

It is well known that several mutation and repair processes preferentially act on single-stranded DNA and, combined with selection pressure, suggest that the mutation patterns within exomes should be different from the genome-wide pattern. This study tests this hypothesis by comparing exome-wide and genome-wide mutation patterns in seven different tumor samples. These seven tumor samples were selected because they contain at least 2000 somatic autosomal single base substitutions (SBSs) within exonic regions and allow a direct comparison with the genome-wide mutation patterns. To determine whether they are statistically the same, 1000 Bootstrap samples, without replacement, were created to generate files with the same number of SBSs as the exome-wide results for each tumor sample. The genome-wide, exome-wide, and 1000 Bootstrap mutation samples were each used to build a somatic autosomal mutation matrix (SAMM), which captures the mutation pattern in the context of the central position of a pentanucleotide. The Manhattan distances between the 1000 Bootstrap SAMMs and the genome-wide SAMM are used to determine whether the difference between the exome-wide and genome-wide SAMMs is a result of a smaller number of SBSs, or if there are slight differences in the mutation patterns. One of the exome-wide SAMMs is statistically the same as its corresponding genome-wide pattern; the others show slight differences. To determine whether the mutation patterns are similar, a distance-dependent 6-nearest neighbor classifier was used to predict the tissue of origin of the exome sample when compared to 908 genome-wide mutation patterns from 12 different tissue types. Six of the seven exome patterns are nearest neighbors to their corresponding genome-wide mutation patterns, and all seven exome patterns are correctly classified as to their tissue of origin. Therefore, even though there may well be differences in specific mutations between the genome and exome, exome mutations still contain the overall mutation pattern of the whole genome and the tissue of origin.