This is a preview of subscription content, access via your institution
Relevant articles
Open Access articles citing this article.
-
ERα/PR crosstalk is altered in the context of the ERα Y537S mutation and contributes to endocrine therapy-resistant tumor proliferation
npj Breast Cancer Open Access 30 November 2023
Access options
Subscribe to this journal
Receive 12 print issues and online access
$259.00 per year
only $21.58 per issue
Buy this article
- Purchase on Springer Link
- Instant access to full article PDF
Prices may be subject to local taxes which are calculated during checkout
References
Schroeder, M.P., Gonzalez-Perez, A. & Lopez-Bigas, N. Genome Med. 5, 9 (2013).
The Cancer Genome Atlas Research Network et al. Nat. Genet. 45, 1113–1120 (2013).
The Cancer Genome Atlas Research Network. Nature 513, 202–209 (2014).
The Cancer Genome Atlas Research Network. Nature 487, 330–337 (2012).
Lawrence, M.S. et al. Nature 505, 495–501 (2014).
Ding, L., Wendl, M.C., McMichael, J.F. & Raphael, B.J. Nat. Rev. Genet. 15, 556–570 (2014).
Acknowledgements
We thank the providers of the data sets used in MAGI, which are listed on the MAGI website. This work is supported by US National Institutes of Health (NIH)/NHGRI grants R01HG005690 and R01HG007069 to B.J.R. B.J.R. is supported by a Career Award at the Scientific Interface from the Burroughs Wellcome Fund, an Alfred P. Sloan Research Fellowship and a US National Science Foundation (NSF) CAREER Award (CCF-1053753). M.D.M.L. is supported by NSF GRFP DGE 0228243. C.C.G. is supported by NSF GRFP DGE 1058262 and NSF IIS-10-16623. The funders had no role in study design, data collection and analysis, decision to publish or preparation of the manuscript.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing financial interests.
Integrated supplementary information
Supplementary Figure 1 Screenshot of the MAGI home page and query interface.
(i) Users select a combination of (public or private) datasets to query. (ii) Users can enter up to 25 genes to query at once. (iii) Alternatively, users can view the mutations in a single sample (Supplementary Fig. 6).
Supplementary Figure 2 Schematic of the software technologies used in MAGI.
MAGI uses D3 client-side to load the mutations and annotations from a web server running Nginx. The web server constructs a JSON object payload using the Node.js framework, which performs the query and collates the resulting data from MongoDB.
Supplementary Figure 3 Mutations in BRAF in glioblastoma tumors in the TCGA Pan-Cancer data set.
Shown is the screenshot of the transcript plot of BRAF mutations in transcript ENST00000288602. Each diamond indicates a missense mutation in an individual sample - five of six are clustered at position 600.
Supplementary Figure 4 Aberrations view of the mutations in the CDKN2A, CDK4 and RB1 genes in the glioblastoma tumors from the TCGA Pan-Cancer data set.
Full ticks represent SNVs while black stripes represent inactivating SNVs, downticks represent deletions, and upticks represent amplifications. Exclusive mutations - those that occur in only one gene in the gene set - are colored blue, and co-occurring mutations are colored orange.
Supplementary Figure 5 Aberrations view of genes in the SWI/SNF complex from the TCGA Pan-Cancer data set.
The genes are enriched for mutations in kidney cancer (KIRC, red) and endometrial cancer (UCEC, brown). Most of the inactivating SNVs and deletions in these genes occur in the BLCA, KIRC, and UCEC cancer types, which is easily seen in MAGI by toggling off the other cancer types in the sample type legend on the right (not shown).
Supplementary Figure 6 An example of a potential missed target of recurrent CNAs by GISTIC2 in TCGA STAD.
Shown are amplifications assigned by GISTIC2 to the FGF19 gene in the TCGA STAD dataset. (a) View of all the amplifications in FGF19. (b) Zoomed in view of the amplifications, whereCCND1 - a well-studied gene frequently amplified in cancer26 - is visible. CCND1 is only 44kb from FGF19, and thus may also be a target of the CNAs in FGF19.
Supplementary Figure 7 Copy-number aberrations in PDGFRA in the TCGA Pan-Cancer data set.
Amplifications in PDGFRA were identified by GISTIC2 to be significantly recurrent in GBM tumors (green bars), although PDGFRA is also amplified in many other cancers. Vertical bar indicates the genomic coordinates of PDGFRA gene.
Supplementary Figure 8 Example of sample linking.
As users scroll over mutations in the transcript view (i), the corresponding samples are highlighted in the aberration matrix (ii), heatmap (iii), and copy number views (iv).
Supplementary Figure 9 Screenshot of MAGI data-upload interface.
(i) Users can choose to upload a single JSON manifest file that includes the URLs for all the data files in a given dataset. (ii) Users select a cancer type for their dataset either from a predefined list of TCGA/ICGC types or by adding their own - ensuring every mutation annotation maps to a particular cancer type. (iii) Users select the data files they want to load into MAGI. MAGI supports six different data types: SNVs, CNAs, other aberrations, heatmap (continuous values), sample annotations, and sample annotation colors. (iv) Users can also upload SNVs in TCGA MAF and CNAs in GISTIC2 format. (v) All data files can be provided as URLs, or uploaded directly to MAGI.
Supplementary Figure 10 Summary of the glioblastoma (GBM) samples from the TCGA Pan-Cancer data set automatically produced by MAGI.
Users can view these summary pages for public datasets or their own uploaded private data. (i) Summary statistics for the mutation data. (ii) Plot of the number of the mutations in each gene in the dataset. The plot is interactive, as users can zoom in and out or scroll over points to get additional information. (iii) Users can also choose from 11 mutation categories for the x- and y-axes (number of SNVs and number of CNAs by default) of the mutations plot. (iv) Sortable and searchable table of the genes in the dataset. (v) Sortable and searchable list of pathways with the most mutations in the dataset. Pathways are from the KEGG and PINdb databases.
Supplementary Figure 11 Screenshot of MAGI annotation interface.
Annotations of (i) genes (ii) expression, (iii) interactions, (iv) mutated residues, and (v) copy number aberrations are viewable as tooltips in each view. MAGI displays annotations interactively as users mouse over different views. Users can also upvote or downvote existing annotations in the tool tips. (vi) MAGI lists the annotations in the database for each gene. (vii) Users can click on individual data points (e.g. interactions in the network view) to pre-load the annotation form on the right.
Supplementary Figure 12 The MAGI gene annotation page shows a table listing all the annotations for a given gene.
The PubMed or PubMed Central ID, the source, and (optionally) the associated cancer, mutation class, mutation type, and protein sequence change are shown for each annotation. Logged-in users can upvote or downvote any of the annotations.
Supplementary Figure 13 MAGI's tumor sample view.
(i) The tumor sample view lists all the aberrations in the given tumor sample in a table. The aberrations are displayed are represented by the gene, mutation class, and locus or protein sequence change, and are ordered by the score of the annotations for that aberration in the MAGI database. The badges next to highlighted genes/classes/loci list the number of annotations for the gene/class/locus. (ii) Users can click on any of the badges to view the annotations in more detail. (iii) The page also includes a table of the attributes (e.g. gender) for the given sample.
Supplementary Figure 14 MAGI interface for computing statistical tests of association between mutation status and sample annotations.
(a) Users view enrichments for their query by following the "Enrichment statistics" link on the view page. (b) The MAGI enrichment statistics page, shown here for the genes PTEN, ERBB2, and EGFR in BRCA samples annotated by the gene expression subtypes from [50]. i) Users choose the sample annotation they want to test from a dropdown of all discrete sample annotation categories. ii) MAGI then cross-classifies samples into a contingency table. iii) Users view enrichments by choosing from a dropdown of statistical tests.
Supplementary Figure 15 Transcript plots for mutations in SMAD2 (top) and SMAD4 (bottom) in the TCGA gastric (STAD; dark green) and Pan-Cancer data sets.
(top) Three of the four mutations in SMAD2 in the TCGA STAD dataset are inactivating nonsense mutations, while the fourth mutation is a missense mutation that occurs in the same location in the MH2 binding domain as two missense mutations in the TCGA colorectal cancer (COADREAD; light green) dataset. (bottom) All of the mutations in SMAD4 in the TCGA gastric dataset occur in the MH2 binding domain. In addition, 12 of the 49 mutations in SMAD4 are mutations at position 361. (Note that only 7 of these mutations are visible in the screenshot.)
Supplementary information
Supplementary Texts and Figures
Supplementary Figures 1–15, Supplementary Methods, Supplementary Note and Supplementary Tables 1–3 (PDF 4375 kb)
Supplementary Software
MAGI software for local installation. See http://magi.cs.brown.edu/ for the most up-to-date version and for the web application (ZIP 2272 kb)
Rights and permissions
About this article
Cite this article
Leiserson, M., Gramazio, C., Hu, J. et al. MAGI: visualization and collaborative annotation of genomic aberrations. Nat Methods 12, 483–484 (2015). https://doi.org/10.1038/nmeth.3412
Published:
Issue Date:
DOI: https://doi.org/10.1038/nmeth.3412
This article is cited by
-
ERα/PR crosstalk is altered in the context of the ERα Y537S mutation and contributes to endocrine therapy-resistant tumor proliferation
npj Breast Cancer (2023)
-
Visualizing and interpreting cancer genomics data via the Xena platform
Nature Biotechnology (2020)
-
Integrative clinical genomics of metastatic cancer
Nature (2017)
