Limits...
MutSpec: a Galaxy toolbox for streamlined analyses of somatic mutation spectra in human and mouse cancer genomes.

Ardin M, Cahais V, Castells X, Bouaoun L, Byrnes G, Herceg Z, Zavadil J, Olivier M - BMC Bioinformatics (2016)

Bottom Line: Results are provided in various formats including rich graphical outputs.An example is presented to illustrate the package functionalities, the straightforward workflow analysis and the richness of the statistics and publication-grade graphics produced by the tool.MutSpec can thus effectively assist in the discovery of complex mutational processes resulting from exogenous and endogenous carcinogenic insults.

View Article: PubMed Central - PubMed

Affiliation: Molecular Mechanisms and Biomarkers Group, International Agency for Research on Cancer, F69372, Lyon, France.

ABSTRACT

Background: The nature of somatic mutations observed in human tumors at single gene or genome-wide levels can reveal information on past carcinogenic exposures and mutational processes contributing to tumor development. While large amounts of sequencing data are being generated, the associated analysis and interpretation of mutation patterns that may reveal clues about the natural history of cancer present complex and challenging tasks that require advanced bioinformatics skills. To make such analyses accessible to a wider community of researchers with no programming expertise, we have developed within the web-based user-friendly platform Galaxy a first-of-its-kind package called MutSpec.

Results: MutSpec includes a set of tools that perform variant annotation and use advanced statistics for the identification of mutation signatures present in cancer genomes and for comparing the obtained signatures with those published in the COSMIC database and other sources. MutSpec offers an accessible framework for building reproducible analysis pipelines, integrating existing methods and scripts developed in-house with publicly available R packages. MutSpec may be used to analyse data from whole-exome, whole-genome or targeted sequencing experiments performed on human or mouse genomes. Results are provided in various formats including rich graphical outputs. An example is presented to illustrate the package functionalities, the straightforward workflow analysis and the richness of the statistics and publication-grade graphics produced by the tool.

Conclusions: MutSpec offers an easy-to-use graphical interface embedded in the popular Galaxy platform that can be used by researchers with limited programming or bioinformatics expertise to analyse mutation signatures present in cancer genomes. MutSpec can thus effectively assist in the discovery of complex mutational processes resulting from exogenous and endogenous carcinogenic insults.

No MeSH data available.


Related in: MedlinePlus

Mutation signatures in Indian OSCC and their suspected origin. Summary results of MutSpec-NMF and MutSpec-Compare analyses obtained on the 106 OSCC samples. a Mutation signatures obtained by NMF with a factorisation value of four. b Comparison of the four OSCC signatures (vertical axis) with 34 reference signatures (horizontal axis) using the cosine similarity method. The heatmap is color-coded according to the cosine value that ranges from 0 to 1. Only reference signatures with a significant match (cosine > 0.9) are labelled. c Number of mutations contributing to each of the four signatures identified, for each sample analyzed. d Average contributions of the four identified signatures to the mutation load of clustered samples and number of samples by cluster
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
getmorefigures.php?uid=PMC4835840&req=5

Fig3: Mutation signatures in Indian OSCC and their suspected origin. Summary results of MutSpec-NMF and MutSpec-Compare analyses obtained on the 106 OSCC samples. a Mutation signatures obtained by NMF with a factorisation value of four. b Comparison of the four OSCC signatures (vertical axis) with 34 reference signatures (horizontal axis) using the cosine similarity method. The heatmap is color-coded according to the cosine value that ranges from 0 to 1. Only reference signatures with a significant match (cosine > 0.9) are labelled. c Number of mutations contributing to each of the four signatures identified, for each sample analyzed. d Average contributions of the four identified signatures to the mutation load of clustered samples and number of samples by cluster

Mentions: To analyse mutation signatures present in the dataset, we ran MutSpec-NMF using the output report of MutSpec-Stat and factorisation value was set to 4. We then compared the obtained signatures with published signatures using the tool MutSpec-Compare. Fig. 3a shows the 4 signatures obtained. Signature A matched best with signature 1 (Age), signature B with signatures 29 (tobacco chewing) and 24 (aflatoxin), signature C with signature 7 (UV) and signature D with signature 13 (APOBEC) (Fig. 3b). MutSpec-NMF also produces a graph showing the total number of SBS per sample and the proportion contributing to the 4 signatures (Fig. 3c). On this graph, one sample is standing out with the largest number of SBS and a close to 100 % contribution to the UV signature (sig 7). Finally, NMF clusters samples based on their signatures composition. From these data, MutSpec-NMF produces a summary analysis that shows the number of samples by cluster and the average contributions of each signature in each cluster (Fig. 3d). In the 106 OSCC samples analysed here, we found one sample likely to be related to UV exposure (high number of SBS corresponding to the previously reported UV signature, sig.7) while a majority of samples (N = 47) had a predominant signature related to tobacco chewing and/or aflatoxin. Another large set of samples (N = 41) had the age signature as the predominant signature, and in a small number of samples (N = 17) the APOBEC signature was the most prominent. Because the cases analysed are from India where tobacco chewing is one major risk factor for oral cancer, it is more likely that the signature B found here is related to tobacco chewing and not aflatoxin. These two signatures (24 and 29) are in fact very close (they share several predominant C > A in specific contexts due to similar mechanisms of the suspected underlying carcinogens) and thus difficult to separate by NMF [5]. The fact that a majority of samples were found to carry this signature confirms the major role of tobacco chewing in the etiology of OSCC in India.Fig. 3


MutSpec: a Galaxy toolbox for streamlined analyses of somatic mutation spectra in human and mouse cancer genomes.

Ardin M, Cahais V, Castells X, Bouaoun L, Byrnes G, Herceg Z, Zavadil J, Olivier M - BMC Bioinformatics (2016)

Mutation signatures in Indian OSCC and their suspected origin. Summary results of MutSpec-NMF and MutSpec-Compare analyses obtained on the 106 OSCC samples. a Mutation signatures obtained by NMF with a factorisation value of four. b Comparison of the four OSCC signatures (vertical axis) with 34 reference signatures (horizontal axis) using the cosine similarity method. The heatmap is color-coded according to the cosine value that ranges from 0 to 1. Only reference signatures with a significant match (cosine > 0.9) are labelled. c Number of mutations contributing to each of the four signatures identified, for each sample analyzed. d Average contributions of the four identified signatures to the mutation load of clustered samples and number of samples by cluster
© Copyright Policy - OpenAccess
Related In: Results  -  Collection

License 1 - License 2
Show All Figures
getmorefigures.php?uid=PMC4835840&req=5

Fig3: Mutation signatures in Indian OSCC and their suspected origin. Summary results of MutSpec-NMF and MutSpec-Compare analyses obtained on the 106 OSCC samples. a Mutation signatures obtained by NMF with a factorisation value of four. b Comparison of the four OSCC signatures (vertical axis) with 34 reference signatures (horizontal axis) using the cosine similarity method. The heatmap is color-coded according to the cosine value that ranges from 0 to 1. Only reference signatures with a significant match (cosine > 0.9) are labelled. c Number of mutations contributing to each of the four signatures identified, for each sample analyzed. d Average contributions of the four identified signatures to the mutation load of clustered samples and number of samples by cluster
Mentions: To analyse mutation signatures present in the dataset, we ran MutSpec-NMF using the output report of MutSpec-Stat and factorisation value was set to 4. We then compared the obtained signatures with published signatures using the tool MutSpec-Compare. Fig. 3a shows the 4 signatures obtained. Signature A matched best with signature 1 (Age), signature B with signatures 29 (tobacco chewing) and 24 (aflatoxin), signature C with signature 7 (UV) and signature D with signature 13 (APOBEC) (Fig. 3b). MutSpec-NMF also produces a graph showing the total number of SBS per sample and the proportion contributing to the 4 signatures (Fig. 3c). On this graph, one sample is standing out with the largest number of SBS and a close to 100 % contribution to the UV signature (sig 7). Finally, NMF clusters samples based on their signatures composition. From these data, MutSpec-NMF produces a summary analysis that shows the number of samples by cluster and the average contributions of each signature in each cluster (Fig. 3d). In the 106 OSCC samples analysed here, we found one sample likely to be related to UV exposure (high number of SBS corresponding to the previously reported UV signature, sig.7) while a majority of samples (N = 47) had a predominant signature related to tobacco chewing and/or aflatoxin. Another large set of samples (N = 41) had the age signature as the predominant signature, and in a small number of samples (N = 17) the APOBEC signature was the most prominent. Because the cases analysed are from India where tobacco chewing is one major risk factor for oral cancer, it is more likely that the signature B found here is related to tobacco chewing and not aflatoxin. These two signatures (24 and 29) are in fact very close (they share several predominant C > A in specific contexts due to similar mechanisms of the suspected underlying carcinogens) and thus difficult to separate by NMF [5]. The fact that a majority of samples were found to carry this signature confirms the major role of tobacco chewing in the etiology of OSCC in India.Fig. 3

Bottom Line: Results are provided in various formats including rich graphical outputs.An example is presented to illustrate the package functionalities, the straightforward workflow analysis and the richness of the statistics and publication-grade graphics produced by the tool.MutSpec can thus effectively assist in the discovery of complex mutational processes resulting from exogenous and endogenous carcinogenic insults.

View Article: PubMed Central - PubMed

Affiliation: Molecular Mechanisms and Biomarkers Group, International Agency for Research on Cancer, F69372, Lyon, France.

ABSTRACT

Background: The nature of somatic mutations observed in human tumors at single gene or genome-wide levels can reveal information on past carcinogenic exposures and mutational processes contributing to tumor development. While large amounts of sequencing data are being generated, the associated analysis and interpretation of mutation patterns that may reveal clues about the natural history of cancer present complex and challenging tasks that require advanced bioinformatics skills. To make such analyses accessible to a wider community of researchers with no programming expertise, we have developed within the web-based user-friendly platform Galaxy a first-of-its-kind package called MutSpec.

Results: MutSpec includes a set of tools that perform variant annotation and use advanced statistics for the identification of mutation signatures present in cancer genomes and for comparing the obtained signatures with those published in the COSMIC database and other sources. MutSpec offers an accessible framework for building reproducible analysis pipelines, integrating existing methods and scripts developed in-house with publicly available R packages. MutSpec may be used to analyse data from whole-exome, whole-genome or targeted sequencing experiments performed on human or mouse genomes. Results are provided in various formats including rich graphical outputs. An example is presented to illustrate the package functionalities, the straightforward workflow analysis and the richness of the statistics and publication-grade graphics produced by the tool.

Conclusions: MutSpec offers an easy-to-use graphical interface embedded in the popular Galaxy platform that can be used by researchers with limited programming or bioinformatics expertise to analyse mutation signatures present in cancer genomes. MutSpec can thus effectively assist in the discovery of complex mutational processes resulting from exogenous and endogenous carcinogenic insults.

No MeSH data available.


Related in: MedlinePlus