Integrate and visualize CH calls

[jjgao]

This issue is for integrating and visualizing clonal hematopoiesis mutations into cBioPortal. From Ptashkin et al, 2018: "Clonal hematopoiesis (CH) is the somatic acquisition of genomic alterations in hematopoietic stem and/or progenitor cells, leading to clonal expansion. In patients with cancer, CH is a common occurrence, associated with aging, smoking, and radiation therapy."

From a technical point of view, we would like to add the CH mutations into the portal so that they can be visualized and analyzed with clinical data and other genomics data.

Pipeline

• [ ] Pull CH calls from CVR
  • hand-deliver MAF to start?
• [ ] Create an MAF: data_mutations_ch.txt
  • use the custom driver annotation for CH-PD (putative driver, based on pathogenicity and allele frequency determined by DMP)
  • should mutation_status be Somatic or Clonal hematopoiesis?
  • Add #sequenced_samples:
• [ ] Create a meta file: meta_mutations_ch.txt with profile type of MUTATION_CH
• [ ] Gene panel?
• [ ] CH-related clinical data (needs more details)?

Backend

• [ ] add profile type MUTATION_CH
• [ ] Importer
  • importing to the same mutation & mutation_event table?
• [ ] Validator

Frontend

• [ ] Add CH mutations into query page (unselected by default)
• [ ] OncoPrint tab: when CH mutations profile is queried, create separate tracks for each query gene (in the gene panel?)
  • Reason for a separate track: CH mutations occurs in non-tumor cells
  • The same colors can be used
  • CH-PD will be displayed as driver
• [ ] Mutations tab: add CH mutations to the table; filter by CH mutations
  • CH-PD should be annotated
• [ ] Study view: Add a new table CH Mutated Genes
• [ ] Patient view: Add CH mutations to the mutations table
  • CH-PD should be annotated

Q: How to deal with sample IDs?

[rptashkin]

@jjgao what is the best way to differentiate between patients with no CH mutations and those that were not analyzed for CH (e.g. no blood sample sequenced by IMPACT)? We can provide a table of samples analyzed if helpful. There will also be cases where multiple blood samples are analyzed from the same patient; these will have unique DMP IDs (e.g. -N02)

[jjgao]

@rptashkin you can add "#sequenced_samples: " to indicate which samples were analyzed for CH, e.g. https://github.com/cBioPortal/cbioportal/blob/master/core/src/test/resources/data_mutations_extended_duplicate_events.txt

Question: are all genes in the impact panel analyzed for CH?

[rptashkin]

@jjgao yes all IMPACT panel genes are analyzed for CH

[jjgao]

Niki, Ahmet, and I had a discussion and decided we should just create a new study for the moment.

@rptashkin please still send us the data :)