Princess Máxima Center for Pediatric Oncology, Heidelberglaan 25, 3584CS, Utrecht, The Netherlands.
Oncode Institute, Jaarbeursplein 6, 3521 AL, Utrecht, The Netherlands.
BMC Genomics. 2022 Feb 15;23(1):134. doi: 10.1186/s12864-022-08357-3.
The collective of somatic mutations in a genome represents a record of mutational processes that have been operative in a cell. These processes can be investigated by extracting relevant mutational patterns from sequencing data.
Here, we present the next version of MutationalPatterns, an R/Bioconductor package, which allows in-depth mutational analysis of catalogues of single and double base substitutions as well as small insertions and deletions. Major features of the package include the possibility to perform regional mutation spectra analyses and the possibility to detect strand asymmetry phenomena, such as lesion segregation. On top of this, the package also contains functions to determine how likely it is that a signature can cause damaging mutations (i.e., mutations that affect protein function). This updated package supports stricter signature refitting on known signatures in order to prevent overfitting. Using simulated mutation matrices containing varied signature contributions, we showed that reliable refitting can be achieved even when only 50 mutations are present per signature. Additionally, we incorporated bootstrapped signature refitting to assess the robustness of the signature analyses. Finally, we applied the package on genome mutation data of cell lines in which we deleted specific DNA repair processes and on large cancer datasets, to show how the package can be used to generate novel biological insights.
This novel version of MutationalPatterns allows for more comprehensive analyses and visualization of mutational patterns in order to study the underlying processes. Ultimately, in-depth mutational analyses may contribute to improved biological insights in mechanisms of mutation accumulation as well as aid cancer diagnostics. MutationalPatterns is freely available at http://bioconductor.org/packages/MutationalPatterns .
基因组中的体细胞突变集合代表了在细胞中起作用的突变过程的记录。可以通过从测序数据中提取相关的突变模式来研究这些过程。
在这里，我们展示了 MutationalPatterns 的下一版本，这是一个 R/Bioconductor 包，它允许对单碱基和双碱基替换以及小插入和缺失的目录进行深入的突变分析。该软件包的主要功能包括进行区域突变谱分析的可能性，以及检测链不对称现象（如损伤分离）的可能性。除此之外，该软件包还包含了确定特征是否可能导致破坏性突变（即影响蛋白质功能的突变）的功能。这个更新的软件包支持更严格的签名拟合，以防止过度拟合。使用包含不同特征贡献的模拟突变矩阵，我们表明即使每个特征只有 50 个突变，也可以实现可靠的拟合。此外，我们还将引导签名拟合纳入其中，以评估签名分析的稳健性。最后，我们将该软件包应用于删除特定 DNA 修复过程的细胞系的基因组突变数据和大型癌症数据集，以展示如何使用该软件包生成新的生物学见解。
这个新版本的 MutationalPatterns 允许更全面的分析和可视化突变模式，以研究潜在的过程。最终，深入的突变分析可能有助于改善对突变积累机制的生物学见解，并有助于癌症诊断。MutationalPatterns 可在 http://bioconductor.org/packages/MutationalPatterns 免费获得。
