Accuracy of mutational signature software on correlated signatures
Abstract Mutational signatures are characteristic patterns of mutations generated by exogenous mutagens or by endogenous mutational processes. Mutational signatures are important for research into DNA damage and repair, aging, cancer biology, genetic toxicology, and epidemiology. Unsupervised learni...
| Main Authors: | , , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Nature Portfolio
2022-01-01
|
| Series: | Scientific Reports |
| Online Access: | https://doi.org/10.1038/s41598-021-04207-6 |
| _version_ | 1798026479435513856 |
|---|---|
| author | Yang Wu Ellora Hui Zhen Chua Alvin Wei Tian Ng Arnoud Boot Steven G. Rozen |
| author_facet | Yang Wu Ellora Hui Zhen Chua Alvin Wei Tian Ng Arnoud Boot Steven G. Rozen |
| author_sort | Yang Wu |
| collection | DOAJ |
| description | Abstract Mutational signatures are characteristic patterns of mutations generated by exogenous mutagens or by endogenous mutational processes. Mutational signatures are important for research into DNA damage and repair, aging, cancer biology, genetic toxicology, and epidemiology. Unsupervised learning can infer mutational signatures from the somatic mutations in large numbers of tumors, and separating correlated signatures is a notable challenge for this task. To investigate which methods can best meet this challenge, we assessed 18 computational methods for inferring mutational signatures on 20 synthetic data sets that incorporated varying degrees of correlated activity of two common mutational signatures. Performance varied widely, and four methods noticeably outperformed the others: hdp (based on hierarchical Dirichlet processes), SigProExtractor (based on multiple non-negative matrix factorizations over resampled data), TCSM (based on an approach used in document topic analysis), and mutSpec.NMF (also based on non-negative matrix factorization). The results underscored the complexities of mutational signature extraction, including the importance and difficulty of determining the correct number of signatures and the importance of hyperparameters. Our findings indicate directions for improvement of the software and show a need for care when interpreting results from any of these methods, including the need for assessing sensitivity of the results to input parameters. |
| first_indexed | 2024-04-11T18:36:06Z |
| format | Article |
| id | doaj.art-1a915049c41a4a2fa8366715e23d7dca |
| institution | Directory Open Access Journal |
| issn | 2045-2322 |
| language | English |
| last_indexed | 2024-04-11T18:36:06Z |
| publishDate | 2022-01-01 |
| publisher | Nature Portfolio |
| record_format | Article |
| series | Scientific Reports |
| spelling | doaj.art-1a915049c41a4a2fa8366715e23d7dca2022-12-22T04:09:14ZengNature PortfolioScientific Reports2045-23222022-01-0112111210.1038/s41598-021-04207-6Accuracy of mutational signature software on correlated signaturesYang Wu0Ellora Hui Zhen Chua1Alvin Wei Tian Ng2Arnoud Boot3Steven G. Rozen4Programme in Cancer and Stem Cell Biology, Duke-NUS Medical SchoolDepartment of Biological Sciences, National University of SingaporeProgramme in Cancer and Stem Cell Biology, Duke-NUS Medical SchoolProgramme in Cancer and Stem Cell Biology, Duke-NUS Medical SchoolProgramme in Cancer and Stem Cell Biology, Duke-NUS Medical SchoolAbstract Mutational signatures are characteristic patterns of mutations generated by exogenous mutagens or by endogenous mutational processes. Mutational signatures are important for research into DNA damage and repair, aging, cancer biology, genetic toxicology, and epidemiology. Unsupervised learning can infer mutational signatures from the somatic mutations in large numbers of tumors, and separating correlated signatures is a notable challenge for this task. To investigate which methods can best meet this challenge, we assessed 18 computational methods for inferring mutational signatures on 20 synthetic data sets that incorporated varying degrees of correlated activity of two common mutational signatures. Performance varied widely, and four methods noticeably outperformed the others: hdp (based on hierarchical Dirichlet processes), SigProExtractor (based on multiple non-negative matrix factorizations over resampled data), TCSM (based on an approach used in document topic analysis), and mutSpec.NMF (also based on non-negative matrix factorization). The results underscored the complexities of mutational signature extraction, including the importance and difficulty of determining the correct number of signatures and the importance of hyperparameters. Our findings indicate directions for improvement of the software and show a need for care when interpreting results from any of these methods, including the need for assessing sensitivity of the results to input parameters.https://doi.org/10.1038/s41598-021-04207-6 |
| spellingShingle | Yang Wu Ellora Hui Zhen Chua Alvin Wei Tian Ng Arnoud Boot Steven G. Rozen Accuracy of mutational signature software on correlated signatures Scientific Reports |
| title | Accuracy of mutational signature software on correlated signatures |
| title_full | Accuracy of mutational signature software on correlated signatures |
| title_fullStr | Accuracy of mutational signature software on correlated signatures |
| title_full_unstemmed | Accuracy of mutational signature software on correlated signatures |
| title_short | Accuracy of mutational signature software on correlated signatures |
| title_sort | accuracy of mutational signature software on correlated signatures |
| url | https://doi.org/10.1038/s41598-021-04207-6 |
| work_keys_str_mv | AT yangwu accuracyofmutationalsignaturesoftwareoncorrelatedsignatures AT ellorahuizhenchua accuracyofmutationalsignaturesoftwareoncorrelatedsignatures AT alvinweitianng accuracyofmutationalsignaturesoftwareoncorrelatedsignatures AT arnoudboot accuracyofmutationalsignaturesoftwareoncorrelatedsignatures AT stevengrozen accuracyofmutationalsignaturesoftwareoncorrelatedsignatures |