February, 2020 Community Call - Heart

Presenters:

  • Michela Noseda - Imperial College London
  • Sara Samari - Imperial College London

Slides:

Video recording:

Summary of the community call

Key learnings

  • The key metadata types for them are clinical, experimental, histology
  • Histology and clinical diagnoses is where it is difficult to integrate between datasets due to different nomenclature, units, frequency etc. that may differ between countries/consortia/clinics
  • Not easy to know up front what metadata might be useful for others, a field that could be available from the clinic but irrelevant for the primary researcher isn’t captured even though it could be useful for other consumers, e.g. allergies
  • Want to capture histostaining of every tissue they collect, currently have this for each region they have analysed but don’t have a good way of sharing
  • Method for defining where image comes from still too vague

Needs from the community

  • Have a standard way of defining histological and clinical metadata that is interoperable across consortia/countries
  • Capture all clinical metadata that is available
  • Method for sharing histological images as well as the metadata that describes where it comes from, and the analysis by a pathologist
  • Better method for defining spatial location of samples
  • Listing of current HCA projects so researchers can see what research is happening and identify collaboration opportunities

Ideas for how to action

  • Characterise how different consortia capture things like duration and frequency of Alcohol use, Tobacco use, Drug use, medications, testing. Define a standard way of capturing that would not lose information.
  • Research whether there are standard ways of capturing diagnosis, is MONDO adequate?
  • Get metadata directly from clinicians/tissue banks and link to data with use of accessions (BioSamples)
  • Encourage uploading of histopathology images along with sequencing data. We can do this but it doesn’t generally happen. Some kind of pilot to show the ability to do this?
  • Implementation of the CCF? (unknown when this will be usable)
  • Resource to enable easier use of ontologies with visualisations
  • Anatomical atlas of the heart?
  • Investigate how to add spatial dimension to ontologies
  • Encourage use of BioStudies for registration. Create a portal that gathers all HCA Biostudies projects - use HCA registry to initially seed it?

Read detailed notes