This mutation portal provides an interactive interface to predictions by CHASMplus. CHASMplus is a machine learning algorithm that discriminates somatic missense mutations as cancer drivers. We summarize results with interactive widgets, provide search capability and an interactive table of results. Visualizations include:
- Mutations on protein structure
- Lollipop diagrams
- Number of driver mutations per gene
- Breakdown of mutation frequency
- Distribution of CHASMplus scores
Search by gene
Enter in the HUGO gene symbol at the top of the page to search for all predicted driver mutations for a particular gene.
All searches will dynamically change the summary plots.
The table provides a listing of all predicted driver mutations meeting your search criteria.
We performed analysis with CHASMplus in two different manners. The first called “pan-cancer” aggregates mutations from multiple cancer types together, while the second was done by analyzing each cancer type separately (“cancer type-specific”). You can select which analysis you want to examine by using the dropdown menu (see Marker #1).
In the case of the pan-cancer results where multiple cancer types contain the same mutation, the listed cancer type has the highest mutation frequency.
You can toggle the display of columns by checking/unchecking boxes (see Marker #2).
You may enter multiple combinations of searches by using the advanced searches available on the table.
- Filter by gene name (exact match, see Marker #3)
- Filter by mutation (fuzzy match, see Marker #4)
- Filter by mutations with at least a certain CHASMplus score (see Marker #5)
- Filter by cancer type (see Marker #6)
- Filter by the mutation frequency of the driver mutation (common, intermediate, or rare, see Marker #7)
Additional variant information:
By clicking on the “More details” link for each variant (see Marker #8), CRAVAT provides more information on the variant, including a lollipop plot, protein structure viewer and gene network graph. You can read help about using these widgets from the CRAVAT and MuPIT help pages.
You can download data in either excel format (.xlsx file) or as a csv file (see Marker #9). You can also download the complete list of TCGA patient IDs used in this analysis here.