癌症至今一直是難解之題，大多的癌症發生是源於一系列的DNA突變，這些突變累積至一定量最終轉變成腫瘤細胞(tumor cell)。造成體細胞突變(somatic mutation)的原因可為細胞內的機制失調或外在環境的致癌因子，如：DNA修復缺陷、DNA胞苷脫氨酶活性、DNA修復機制失調、馬兜鈴酸(aflatoxin)等。這些致癌因子造成DNA突變，而體細胞在不斷的修復與累積突變的過程中會留下損傷，這些傷疤被稱為「突變特徵」。研究顯示這些因DNA突變所留下來的突變特徵對應於內在或外在的致癌因子。因此，解析出位於癌症基因體上的突變特徵可以幫助癌症疾病的治療。在2015年時，維康桑格研究所(Wellcome Trust Sanger, WTSI)建立了從40種源於美國癌症基因體圖譜計畫(TCGA)和國際癌症基因組聯盟(ICGC)的癌症分解出30種不同的突變特徵，並收入於COSMIC資料庫中，且其中半數的突變特徵擁有相對應的致癌因子，如：菸草、紫外線、DNA編輯酶APOBEC胞嘧啶脫氨酶突變等。在今年2019五月時，WTSI發表了新的突變特徵資料庫(COSMICv3)，此突變特徵資料庫相較於2015年合併了從19,148全外顯子Pan-Cancer Analysis of Whole Genomes 分解出共95種突變特徵，分別為67種單鹼基取代特徵(single base substitution signatures), 17種雙鹼基取代特徵(doublet base substitution signatures)和11種小片段插入和缺失特徵(small insertion and deletion)。目前市面上多數分析突變特徵的工具並未收入2019年的95種突變特徵，並侷限於熟悉程式碼編輯的資訊專業人士。因此，我們參考2018的mSignatureDB，架構資料更完整的mSignautreDB 其收錄95種突變特徵(SBS/DBS/ID) 橫跨84種源於ICGC/TCGA癌症，提供一個完整的資料視覺化資料庫，使更多研究者能投身於此突變特徵研究項目。

Cancer is originated from somatic mutations which reflect the consequence of changes in DNA through multiple mutational processes during tumor development. Mutational signatures have been proved to provide a valuable pattern to assist to figure out the causative mutagens, such as infidelity of DNA polymerase, defective DNA mismatch repair, and aflatoxin. In 2015, the Wellcome Trust Sanger Institute (WTSI) have extracted 30 distinct mutational signatures from over 10,000 cancer exomes across 40 human cancer types based on TCGA/ICGC and include in COSMIC. Also, it indicates that half of mutational signatures are associated to the extrinsic or intrinsic carcinogenic factors, for example, the enzymatic activity of DNA cytidine deaminases (APOBECs), the deficiency of DNA mismatch repair, mutations in POLE, tobacco, and ultraviolet. In 2019, WTSI established COSMICv3 which identified 67 single base substitution signatures (SBS), 17 doublet base substitution signatures (DBS), and 11 small insertion and deletion (ID) signatures from 19,148 whole exomes derived from Pan-Cancer Analysis of Whole Genomes (PCAWG), providing unambiguously mutational signatures. Up to now, numerous bioinformatic packages have been developed to decipher mutational signatures from tumor genomes, yet limited to a small group of bioinformatics experts. As a result, based on database we have constructed in 2018, we established a more comprehensive database, mSignatureDB, which integrated with 95 mutational signatures from 84 cancer projects in TCGA/ICGC to provide an online database with valuable visualization module and complete mutational signatures.

Acknowledgements.......4
Abbreviations...5
摘要....6
Abstract........7
1. Introduction....8
2. Materials and Methods...13
2.1 Dataset....13
2.1.1 84 cancer projects....14
2.1.2 50 oral cancer samples....14
2.2 Non-negative matrix factorization (NMF) algorithm...15
2.3 Mutation Files......16
2.4 Compendium of Mutational Signatures in the Catalogue of Somatic Mutations in Cancer (COSMIC).....17
2.5 Computation Environments.....18
2.6 Mutational Signatures Analysis......18
2.6.1 Refitting Signatures Analysis.....19
2.6.2 De Novo Signatures Analysis.......20
3. Results and Discussions......21
3.1 Framework of mSignatureDB....21
3.2 Browse......21
3.3 Search......22
3.4 Analysis....24
3.5 Example of use – 50 oral cancers....25
3.6 Comparison to mSignatureDB in 2018...26
Tables...28
Table 1. Comparison of signature analysis tools..28
Table 2. 84 cancer projects from ICGC, TCGA and PCAWG...30
Figures...33
Figure 1. Framework of non-negative matrix factorization (NMF) algorithm......33
Figure 2. An example of partial VCF file........33
Figure 3. Mutational Signature workflow..34
Figure 4. Structure of mSignatureDB.....35
Figure 5. Analysis page.........36
Figure 6. The workflow of deciphering mutational signatures.....37
Figure 7. Browse page.....39
Figure 8. Somatic mutations in project...43
Figure 9. Barplot of different mutational types...44
Figure 10. The distributions of mutational signatures in each sample....46
Figure 11. Cross-project comparison with SBS signatures...48
Figure 12. Mutation counts.....49
Figure 13. Landscape of signatures......50
Figure 14. The estimation plot....51
Figure 15. De novo signatures barplot....52
Figure 16. The contributions of de novo signatures to individual samples...52
Reference.......53
Appendix........56

1. Olivier, M., M. Hollstein, and P. Hainaut, TP53 mutations in human cancers: origins, consequences, and clinical use. Cold Spring Harb Perspect Biol, 2010. 2(1): p. a001008.
2. Hollstein, M., et al., New approaches to understanding p53 gene tumor mutation spectra. Mutat Res, 1999. 431(2): p. 199-209.
3. Kucab, J.E., et al., A Compendium of Mutational Signatures of Environmental Agents. Cell, 2019. 177(4): p. 821-836 e16.
4. Hainaut, P., M. Olivier, and G.P. Pfeifer, TP53 mutation spectrum in lung cancers and mutagenic signature of components of tobacco smoke: lessons from the IARC TP53 mutation database. Mutagenesis, 2001. 16(6): p. 551-3; author reply 555-6.
5. Pfeifer, G.P., p53 mutational spectra and the role of methylated CpG sequences. Mutat Res, 2000. 450(1-2): p. 155-66.
6. Brash, D.E., et al., A role for sunlight in skin cancer: UV-induced p53 mutations in squamous cell carcinoma. Proc Natl Acad Sci U S A, 1991. 88(22): p. 10124-8.
7. Witkin, E.M., Ultraviolet-induced mutation and DNA repair. Annu Rev Microbiol, 1969. 23: p. 487-514.
8. Giglia-Mari, G. and A. Sarasin, TP53 mutations in human skin cancers. Hum Mutat, 2003. 21(3): p. 217-28.
9. Pena-Diaz, J., et al., Noncanonical Mismatch Repair as a Source of Genomic Instability in Human Cells. Mol Cell, 2017. 67(1): p. 162.
10. Pilati, C., et al., Mutational signature analysis identifies MUTYH deficiency in colorectal cancers and adrenocortical carcinomas. J Pathol, 2017. 242(1): p. 10-15.
11. Pfeifer, G.P., et al., Tobacco smoke carcinogens, DNA damage and p53 mutations in smoking-associated cancers. Oncogene, 2002. 21(48): p. 7435-51.
12. Alexandrov, L.B., et al., Mutational signatures associated with tobacco smoking in human cancer. Science, 2016. 354(6312): p. 618-622.
13. Pfeifer, G.P., Y.H. You, and A. Besaratinia, Mutations induced by ultraviolet light. Mutat Res, 2005. 571(1-2): p. 19-31.
14. Harris, R.S., Cancer mutation signatures, DNA damage mechanisms, and potential clinical implications. Genome Med, 2013. 5(9): p. 87.
15. Alexandrov, L.B., et al., A mutational signature in gastric cancer suggests therapeutic strategies. Nat Commun, 2015. 6: p. 8683.
16. Stratton, M.R., P.J. Campbell, and P.A. Futreal, The cancer genome. Nature, 2009. 458(7239): p. 719-24.
17. Alexandrov, L.B., et al., Deciphering signatures of mutational processes operative in human cancer. Cell Rep, 2013. 3(1): p. 246-59.
18. Nik-Zainal, S., et al., Mutational processes molding the genomes of 21 breast cancers. Cell, 2012. 149(5): p. 979-93.
19. Alexandrov, L.B., et al., Signatures of mutational processes in human cancer. Nature, 2013. 500(7463): p. 415-21.
20. Helleday, T., S. Eshtad, and S. Nik-Zainal, Mechanisms underlying mutational signatures in human cancers. Nat Rev Genet, 2014. 15(9): p. 585-98.
21. Alexandrov, L.B. and M.R. Stratton, Mutational signatures: the patterns of somatic mutations hidden in cancer genomes. Curr Opin Genet Dev, 2014. 24: p. 52-60.
22. Forbes, S.A., et al., COSMIC: exploring the world's knowledge of somatic mutations in human cancer. Nucleic Acids Res, 2015. 43(Database issue): p. D805-11.
23. International Cancer Genome, C., et al., International network of cancer genome projects. Nature, 2010. 464(7291): p. 993-8.
24. Omichessan, H., G. Severi, and V. Perduca, Computational tools to detect signatures of mutational processes in DNA from tumours: A review and empirical comparison of performance. PLoS One, 2019. 14(9): p. e0221235.
25. Lee, D.D. and H.S. Seung, Learning the parts of objects by non-negative matrix factorization. Nature, 1999. 401(6755): p. 788-91.
26. Joliffe, I.T. and B.J. Morgan, Principal component analysis and exploratory factor analysis. Stat Methods Med Res, 1992. 1(1): p. 69-95.
27. Rosenthal, R., et al., DeconstructSigs: delineating mutational processes in single tumors distinguishes DNA repair deficiencies and patterns of carcinoma evolution. Genome Biol, 2016. 17: p. 31.
28. Maura, F., et al., A practical guide for mutational signature analysis in hematological malignancies. Nat Commun, 2019. 10(1): p. 2969.
29. Fischer, A., et al., EMu: probabilistic inference of mutational processes and their localization in the cancer genome. Genome Biol, 2013. 14(4): p. R39.
30. Gehring, J.S., et al., SomaticSignatures: inferring mutational signatures from single-nucleotide variants. Bioinformatics, 2015. 31(22): p. 3673-5.
31. Alexandrov, L.B., et al., The Repertoire of Mutational Signatures in Human Cancer. 2019: p. 322859.
32. Gori, K. and A. Baez-Ortega, sigfit: flexible Bayesian inference of mutational signatures. bioRxiv, 2018.
33. Blokzijl, F., et al., MutationalPatterns: comprehensive genome-wide analysis of mutational processes. Genome Med, 2018. 10(1): p. 33.
34. Huang, P.J., et al., mSignatureDB: a database for deciphering mutational signatures in human cancers. Nucleic Acids Res, 2018. 46(D1): p. D964-D970.
35. Petljak, M., et al., Characterizing Mutational Signatures in Human Cancer Cell Lines Reveals Episodic APOBEC Mutagenesis. Cell, 2019. 176(6): p. 1282-1294 e20.
36. Bergstrom, E.N., et al., SigProfilerMatrixGenerator: a tool for visualizing and exploring patterns of small mutational events. BMC Genomics, 2019. 20(1): p. 685.
37. Chen, T.W., et al., APOBEC3A is an oral cancer prognostic biomarker in Taiwanese carriers of an APOBEC deletion polymorphism. Nat Commun, 2017. 8(1): p. 465.
38. Siva, N., 1000 Genomes project. Nat Biotechnol, 2008. 26(3): p. 256.
39. Gao, Z., et al., Interpreting the Dependence of Mutation Rates on Age and Time. PLoS Biol, 2016. 14(1): p. e1002355.
40. Zhang, Z. and M. Gerstein, Patterns of nucleotide substitution, insertion and deletion in the human genome inferred from pseudogenes. Nucleic Acids Res, 2003. 31(18): p. 5338-48.
41. Jee, S.H., et al., Smoking and cancer risk in Korean men and women. Cancer Causes Control, 2004. 15(4): p. 341-8.
42. Lin, C.J., Projected gradient methods for nonnegative matrix factorization. Neural Comput, 2007. 19(10): p. 2756-79.