How the Chan Zuckerberg Science Initiative plans to solve disease by 2100

Is anyone on here involved with the HCA by any chance? I have a question about your efforts.

When I was working in neuroimaging, one of the greatest hindrances to effective open data sharing was the lack of a standard for how data should be organized. Basically, what this amounted to was people from different labs would just throw their data on hard drives and ship them out to people who requested it as-is, with a variety of different organizational schemes (or no organizational scheme at all / raw data directly from a scanner with lots of missing and somewhat important metadata). There was an effort to change that (http://bids.neuroimaging.io/) although I'm not sure how well it has caught on. I've had discussions with friends in other fields (bioinformatics) where apparently there really isn't any standard at all and data organization is basically done according to the researcher's (or postdoc/grad student's) own idiosyncratic methods.

I've always felt like this sort of "here's the dataset as-is, take it or leave it" distribution severely limits the discoverability of open data, since it makes it substantially harder to index and make it easy to search using a variety of relevant criteria (in the case of neuroimaging, it'd be nice to be able to narrow down the kinds of brain images out there based on scanning modality or scan sequence). It also hurts the potential re-usability of datasets, since distributing data without properly populating key metadata fields can make it effectively useless in many cases.

With this context, I was wondering what efforts the HCA folks are making towards dataset standardization. Maybe this is all still in the works / a little to early to ask, but how is the data ingestion phase described on the HCA website [1] going to conform data and ensure that data is usable?

[1] https://www.humancellatlas.org/data-sharing

tl;dr By re-applying lessons learned in mapping biological functions to genetic origins in c. elegans, one idea is to use computational biology, and inventory all human cell types, to accurately trace and debug human physiology at a molecular level. Through a culture of open creativity, other ideas are also welcome.