Spatially resolved multiomics of human cardiac niches

Wkt8 commented 1 year ago

[ ] Convert published metadata to HCA spreadsheet
[ ] Manually curate dataset to meet HCA metadata standard
[ ] Collect any matrix and cell-type annotation files
[ ] Are the analysis files suitable for CellxGene? If something is missing get in touch with the authors to request it
[ ] Upload sheet to validate metadata
[ ] Transfer raw files to ingest to validate data files
[ ] Check linking using ingest graph validator
[ ] Ask the Secondary Wrangler for an end-to-end review of the project. Ask the Expertise Wrangler to review specific tabs if needed
[ ] Submit dataset to Production
[ ] Complete the Export SOP
[ ] Convert project data to SCEA format following the SCEA conversion SOP if appropriate

Wkt8 commented 1 year ago

Spreadsheet complete - next step is to try and submit it to ingest for validation.

arschat commented 11 months ago

Nice work on a huge project, Wei!

Here are my comments:

Donor

D1 sex in supplementary table 1 is female while on spreadsheet is unknown.
Some donors (D1, D2, D4, D11) do not have a linking with specimens I did some more digging and these donors exist in the h5ad, but their sangerID points to the previous ENA project, that has been wrangled before but metadata are not available on DCP. In ingest the submission is archived with many metadata errors. I am not sure what is the case here, but for the sake of this secondary review, the aforementioned donors have been wrangled on another published DCP project. We could either omit the donors, or add the existing values in this project too.
For all donors, biomaterial name has an excel function not a value
For donors that no diseases is recorded, we could add normal for known diseases

imaged_specimen.slice_thickness in FFPE samples is noted as 10 μm while in protocol and paper is noted as 5 μm

fastq files from Visium, although the input biomaterial is in the cell_suspension input column, they are referenced as imaged_specimen (i.e. HCAHeartST9383358_imaged_specimen) which would be in separate imaged_specimen column

Tiff image files for all imaged specimens could be added from heartcellatlas.org in the 3rd dropdown menu (Visium - Interactive sections with cell2location prediction minimal abundances (q05cell_abundance))

log-normalized h5ads do not contain ATAC data. There is a separate h5ad in the heartcellatlas.org site (download here) with the ATAC Peak matrix.

arschat commented 7 months ago

Project was stuck in exporting and was later re-triggered manually by Enrique. Discussion here.

On ingest it is still seems stuck in exporting, but probably can be changed to exported.

arschat commented 6 months ago

State changed (tracker re-started), verified in browser.