The Repertoire of Mutational Signatures in Human Cancer

Abstract

40 Somatic mutations in cancer genomes are caused by multiple mutational processes each of 41 which generates a characteristic mutational signature. Using 84,729,690 somatic mutations 42 from 4,645 whole cancer genome and 19,184 exome sequences encompassing most cancer 43 types we characterised 49 single base substitution, 11 doublet base substitution, four 44 clustered base substitution, and 17 small insertion and deletion mutational signatures. The 45 substantial dataset size compared to previous analyses enabled discovery of new signatures, 46 separation of overlapping signatures and decomposition of signatures into components that 47 may represent associated, but distinct, DNA damage, repair and/or replication mechanisms. 48 Estimation of the contribution of each signature to the mutational catalogues of individual 49 cancer genomes revealed associations with exogenous and endogenous exposures and 50 defective DNA maintenance processes. However, many signatures are of unknown cause. 51 This analysis provides a systematic perspective on the repertoire of mutational processes 52 contributing to the development of human cancer including a comprehensive reference set 53 of mutational signatures in human cancer. 54 55 56

Characterizing Mutational Signatures in Human Cancer Cell Lines Reveals Episodic APOBEC Mutagenesis

“…Most mutations caused by APOBEC mutagenesis during in vitro culture were approximately evenly distributed over the genome, recapitulating the pattern generally observed in cancers in vivo (Figure 5A) (Alexandrov et al., 2018). Foci of localized APOBEC-associated hypermutation (Nik-Zainal et al., 2012a, Roberts et al., 2012), kataegis , were also acquired during in vitro culture of some cell lines (Figure 5) predominantly occurring in clones with genome-wide SBS2 and SBS13 and with more foci in samples with higher rates of genome-wide mutagenesis (Figure 5).…”

“…A small subset of signatures was absent from the examined datasets (SBS7c, SBS12, SBS16, SBS24) or found less often than expected (e.g., SBS3) (Alexandrov et al., 2018). These may be due to the small numbers of somatic mutations in exome sequences, the small numbers of mutations some signatures contribute to individual cancers, the obscuring presence of residual germline variants, the relatively featureless profiles of some signatures that may be more difficult to detect, and/or the genuine absence of the signatures (Alexandrov et al., 2013b).…”

“…For patterns of mutational signatures, see Figure S1. The figure format follows the annotation of mutational signatures across a large set of primary human cancers done previously (Alexandrov et al., 2018). We thank the members of the International Cancer Genome Consortium (ICGC) Pan-Cancer Analysis of Whole Genomes (PCAWG) project for the figure design.

Figure S1Core Set of the Annotated Mutational Signatures, Related to Figures 1, 3, 5, and 6(A) The core set of the mutational signatures, including the Platinum set of the PCAWG signatures and SBS25 discovered in Hodgkin’s lymphoma cell lines.

…”

“…Any signature violating a biologically meaningful constraint (Alexandrov et al., 2018) based on a transcriptional strand bias or a total number of somatic mutations was excluded from the set $\text{Q}$. Further, for the 1,001 cell lines any signature violating a biologically meaningful constraints (Alexandrov et al., 2018) based on number of indels was also excluded from the set $\text{Q}$. Further, any $\overrightarrow{S_r}\times E_r$ for which the cosine similarity between$\hat{C}$ and $\overrightarrow{C}$ does not increase more than 0.01 are removed, where $\hat{C}={\sum }_{r=1}^q\left(\stackrel{true\to }{S_r}\times E_r\right)$.…”

Pathway and network analysis of more than 2,500 whole cancer genomes

“…We performed a permutation-based enrichment test for mutation signatures from PCAWG mutation signatures analysis 18 . We identified the most likely mutation signature for each non-coding mutation in PID-N genes and compared them to randomly chosen non-coding mutations in non-PID-N genes.…”