GSE132065 - GSE132065_BloodTimeHuman

[ipediez]

Project short name:

GSE132065_BloodTimeHuman

Primary Wrangler: @ipediez

Enrique

Secondary Wrangler:

Associated files

• Google Drive: https://drive.google.com/drive/folders/1LuIdzCrdUig-aFgSZuHy6mGfCDNTt5VB?usp=sharing

Published study links

• Paper: Sampling time-dependent artifacts in single-cell genomics studies

• Accessioned data: GSE132065

Ingest

Key Events

• [x] Convert published metadata to HCA spreadsheet
• [x] Manually curate dataset to meet HCA metadata standard
• [x] Collect any matrix and cell-type annotation files
• [x] Upload sheet to validate metadata
• [x] Check linking using ingest graph validator
• [x] Transfer raw files to ingest to validate data files
• [x] Ask the Secondary Wrangler for an end-to-end review of the project. Ask the Expertise Wrangler to review specific tabs if needed
• [x] Submit dataset to Production
• [x] Complete the Export SOP
• [ ] Convert project data to SCEA format following the SCEA conversion SOP if appropriate

[ipediez]

Some files are not downloading from ENA through move_data_from indsc.py and I'm having trouble downloading them via wget. I have tried:

• wget
• the move_data_from_insdc.py script
• @ESapenaVentura tried through Aspera
• SRA Cloud delivery

None of the methods above seems to work, even the cloud delivery is yielding an error. I have sent an email to SRA asking about the error.

[ipediez]

I'm waiting for SRA to solve their problem with cloud delivery, as it seems to be not working for anyone. Meanwhile, I'm trying to get via wget the files that I'm missing.

[ipediez]

Ready for secondary review by @ESapenaVentura

[gabsie]

will be reviewed todayby @ESapenaVentura

[ipediez]

@ESapenaVentura to work on this today

[jacobwindsor]

@ESapenaVentura reviewed and @ipediez to make changes

[ESapenaVentura]

I have reviewed the dataset! A couple of notes:

Donor organism

• If you could link the donor to the ID given in the study it would be great (Not a must, but may help scientists looking at the matrices + our metadata)

Collection protocol

• Description: Please correct the format of the protocol description (There are newlines in weird places)
• pbmc activation protocol: Please correct the description to add the 3 females (Or make it so that it doesn’t state the amount of donors, since the metadata in the paper is inconsistent with what you found in the RDS file)

Specimen

• Organ part: For the specimens that were extracted with the non-cll collection protocol, please add “venous blood” (UBERON:0013756) as organ part

 Enrichment protocol

For the cell lines that had T-Cell activation, maybe it’s worth to add the activation as an enrichment protocol?

For T cell activation, cells were thawed in MACS buffer (1X PBS, 4% FBS, 2 mM EDTA), centrifuged during 5 min at 700×g and RT, and resuspended in pre-warmed culture media (RPMI, 1% Pyruvate, 20% FBS, Pen/Strep, DNase 100 U/ml). A TC20™ automated cell counter was used to assess cell number and viability. The number of only viable cells was used to calculate volumes for cell seeding. For each condition, 200,000 live cells were seeded into two wells of a 96-well round bottom plate (Sigma Aldrich) for a total of 400,000 cells per condition (time point). Dynabeads Human T-Activator CD3/CD28 (Thermo Fisher Scientific) were transferred to a 1.5-ml tube (5 μl/well), washed twice with 1 ml of cell culture media, and resuspended with 10 volumes of cell culture media. Fifty microliters of resuspended beads was added to each well for T cell activation and expansion. Cells were incubated during 24 h at 37 °C with 5% CO2 and 5% humidity. The remaining cells (~ 350,000 cells per condition) were used as a control (day 0) for T cell activation. Cells subjected to T cell activation protocol were collected in a 1.5-ml tube and stained with DAPI (Thermo Fisher Scientific) at 1 μM final concentration. DAPI-negative live individual cells were sorted with a BD FACSAria™ Fusion Flow cytometer (BD Biosciences) in 1X PBS supplemented with 0.05% BSA.

Not super sure where the T-Cell activation would fall though, but I think it can be considered an enrichment protocol?

Cell line

• Cell type field: I would add “T-Cell” here
• Tissue: I would add “blood” here
• Disease: These were extracted from patients with leukaemia, so I’d add the disease that the donors have
• Missing Genus species

Cell suspension:

• Similar to cell line, I’d add “T-Cell” to the selected cell type(s) field for the cell suspensions coming from the CLL patients
• For the SS2 samples, add “1” to estimated cell count (We always do this with SS2)
• For the 10X samples, please add “5000” to estimated cell count (In the paper it states "We set the --chemistry and --expect-cells flags of cellranger count to “SC3Pv3” and “5000”, respectively.”)

Library preparation protocol

• Should probably create a separate protocol for HTO library prep. They are a bit mixed up but I’d create that to specify that HTO libraries were prepared. I’ve been looking into it and I feel they did it for cell surface protein profiling (https://www.ebi.ac.uk/ols/ontologies/efo/terms?iri=http%3A%2F%2Fwww.ebi.ac.uk%2Fefo%2FEFO_0030010&viewMode=All&siblings=false#), and you could link both library preparation methods to the data that contains HTO libraries. Happy to discuss about that if you have any doubts
• Missing primer in ATAC-Seq

Analysis protocol

• Missing analysis protocol tab

[ipediez]

New term requested: T cell activation

[ipediez]

The new term is EFO:0030037 (T cell activation assay). It will be in HCAO following the release scheduled for Tue 16th.

[ESapenaVentura]

I have reviewed the dataset again, this time to the very end :) First of all, congrats, this is a very big dataset and I think you've done an outstanding job on describing the experimental design!

I have just a couple of comments and corrections:

Enrichment protocol

• Added facs sorting and linked it in Cell line enrichment protocols

Cell line

• Missing ontology and ontology_label for tissue
• Missing “disease” field (Fill normal if no info)

Cell suspension

• Is T-cell as selected cell type correct for all cell suspensions? Or only for those ones that come from a cell line?

Sequence file

• Added HTO library prep to the fastq HTO libraries (Please check that I haven’t missed any)
• I noticed some fastq (t-cell lines) link to multiple cell suspensions, but if you go to those runs (e.g.https://www.ncbi.nlm.nih.gov/sra/?term=SRR11614695)), you can see at the bottom that only 1 sample was linked. I think each of those samples can be linked to a single run

[gabsie]

@ipediez still needs to upload the files.

[ipediez]

Graph validator status: pending since 12:00

[ipediez]

Graph validator yields three errors that shouldn't be there:

• contains_umi_barcode_info.adoc . This is related with ingest graph validator issue 49, as this dataset also contains scATAC-seq
• ensure_seq_libr_prep_protocols.adoc . This is related with files with two associated library preparation protocols, which should be allowed.
• paired_end_2_files.adoc . I get this error at ingest, but not at EC2, so I don't know where it's coming from. I've checked the test code and it could be related with the files with two associated library preparation protocols.

The last two errors are related with: https://github.com/ebi-ait/ingest-graph-validator/issues/50

[ESapenaVentura]

About the errors, I have discovered that they will be solved once the changes in the error messages are deployed into production, except contains_umi_barcode_info.adoc, which will need to be addressed

[ESapenaVentura]

contains_umi_barcode_info.adoc fix: https://github.com/ebi-ait/ingest-graph-validator/pull/51

[gabsie]

hey @ipediez and @ESapenaVentura in your absence, data browser team has requested that somebody sense checks this project in the browser: https://data.humancellatlas.org/explore/projects/5b328561-4a97-40ac-b7ad-6a90fc59d374?catalog=dcp12

they are surprised by the 3-level subgraph structure. can you confirm all looks good?

[ESapenaVentura]

There was a discussion around this dataset - Moving to needs update and needs some further discussion

[Wkt8]

Update was based on whether or not specimens could come from specimens. Data is not technically incorrect, but does not conform to how UCSC and us want to wrangle this. Not advised to do this manually. Either - new submission via bulk updates with deletion of entities. Useful to check if linkings can be changed via bulk updates. Or - soft delete of project with a reupload using the same project uuid.

• check if linkings can be updated via bulk updates
• check how easy / difficult it would be to delete entities for this submission

[Wkt8]

Check with alegria as she has done this before. Communicate on the dcp-ops channel about deleting biomaterials, and together with a dev to delete entities from the staging area.

[MightyAx]

Previously submitted entities can't be deleted but they can be ignored / orphaned.

[Wkt8]

Waiting for @ESapenaVentura to return before we discuss this further.

[ofanobilbao]

@ESapenaVentura @Wkt8 did we reach any conclusions on this?

[ESapenaVentura]

Waiting on implementation on delta staging areas because deletion of entities is necessary

[ofanobilbao]

https://app.zenhub.com/workspaces/operations-5fa2d8f2df78bb000f7fb2b5/issues/ebi-ait/hca-ebi-wrangler-central/912 - Is blocking any updates to this project in Ingest, including adding the info in this ticket in order to close this ticket while we wait for this functionality

[ofanobilbao]

I have managed to add the required notes (including reference to this ticket in Ingest back office notes) so I am closing this until further notice

[Wkt8]

Hannes has commented that there are fastqs listed as analysis_file entities which are breaking processes in Azul ticket

[idazucchi]

checking if this project can be updated to correct the fastq file issue or if it would require a soft deletion this could be a chance to correct the other modeling issues with this project (specimen to specimen linking, fastq files as input for analysis files)

[ESapenaVentura]

There is an issue (Defined here) regarding the UMI files being described as analysis files.

The main issue here is that the UMI fastq files were defined as analysis files, which is breaking the browser.

I am investigating what are our options

[ESapenaVentura]

Regarding the issue in the previous comment:

• Found the issue: it's a combination of atac-seq is complicated + poor labelling from the authors
• Solution outwards is a soft deletion. There are more solutions, but they will be very complicated and I don't think we'll get it quickly (if we push for it, and we don't have capacity to take another push), resulting in the project being deleted from the portal for a looong time.
• Possible solutions in ingest:
  • Delete the analysis_files (6) and create new sequence_files (6), delete analysis protocol and replace it in process for library prep protocol. Re-wire the whole set of new fastq files. BIG ISSUE: The upload API is a mistery, and creating/deleting files is not as easy as it sounds - I forecast a lot of issues doing it.
  • Download the current spreadsheet and use it alongside the script I wrote to create a new submission that maintains UUIDs, but changes the analysis_files to sequence_files and the changes above to the protocols.

Based on what we decide it's the best solution, we can start working on the problem