HumanLiverImmuneCells_GSE125188

[ami-day]

Dataset/group this task is for:

HumanLiverImmuneCells GEO accession: GSE125188

Wrangler responsible for this dataset/lab:

Primary wrangler: Ami

File location

https://docs.google.com/spreadsheets/d/1jXH4bIpls-birQicpF8JZJbFU8TLYiYrCuuBH8Dqizs/edit#gid=1240201541

Acceptance criteria for the task:

• [x] Convert metadata

• [x] Add ontologies

• [x] Assign secondary reviewer and make any necessary changes

• [x] Validate linking with graph validator

• [x] Validate metadata: upload to ingest staging

• [x] Obtain cell type annotations and cellxgene matrix (either from publication supplements or by contacting the authors)

• [x] Complete submission: upload metadata and fastq files in ingest production

[mshadbolt]

Changes I made in the MS version of the spreadsheet

• For Project role for wranglers we need to use the text Human Cell Atlas wrangler as this is used by the browser I believe
• added information from the supplementary file of the paper, i.e. sex, age, cause of death, death type
• changed alive at collection to 'no'
• I think I would change 'liver transplant' to a different term like collecting specimen from organ postmortem EFO:0009627 or dissection EFO:0003856
• I also wouldn't use the term 'biopsy' for donors that are no longer living, not sure if that is my own bias, but I think of biopsy more from living donors
• The GEO seems to weirdly have 9 'samples' whereas the spreadsheet has 8, i think the spleen sample is missing from donor 3

I saved a version of the spreadsheet with my initials at the end in the google drive. Would also be good to save the matrix and cell types into the google drive folder.

[ami-day]

Thanks for the review @mshadbolt

There is some confusion about the sample donors: I understood samples were taken from living donors who were receiving a transplant, whereas you have understood it to be deceased organ donors annotated in the supplements. I am going to contact the authors to ask. It might also be a mix: PBMCs from the living donors and liver and spleen from the deceased donors.

About this comment: 'The GEO seems to weirdly have 9 'samples' whereas the spreadsheet has 8, i think the spleen sample is missing from donor 3'. This is because when I transferred the fastq from the ncbi s3 bucket they were not available for the donor 3 spleen sample; I removed this sample from the metadata sheet.

Other than that I agree with the changes you have made and hopefully the authors will get back if I message them. Will also try to get matrices and cell annotations. Thanks.

[ami-day]

Emailed the authors and waiting for a response.

[ami-day]

@mshadbolt I just looked through your review comments again and the paper supplements, and realise you are correct that the samples were from deceased organ donors. I had missed the donor organism info. at the end of the supplements the first time around. Therefore, I have kept the changes you indicate above. I also changed 'liver transplant' and 'biopsy' in the collection protocol tab to 'collecting specimen from organ postmortem' EFO:0009627 since I can now see the 'DBD' in the supplements.

I will email the authors again today to ask about the missing spleen sample files and if they don't get back by mid next week, I'll continue to upload this dataset excluding this particular sample.

[ami-day]

The authors got back and the above issues have been resolved. The metadata is valid in ingest prod. and currently syncing the fastq files.

[ami-day]

Converting to SCEA: E-HCAD-32

[ami-day]

Validated and uploaded to scea gitlab