hepaticSpatialProteogenomics - Guilliams

[ESapenaVentura]

Project short name:

hepaticSpatialProteogenomics

Primary Wrangler:

Irene Perez-Diez

Secondary Wrangler:

Ida

Associated files

• Google Drive: https://drive.google.com/drive/folders/1g0h738T-dY9VQ-GQrx9OWPSJdzO5jkHv
• Ingest: https://contribute.data.humancellatlas.org/projects/detail?uuid=425c2759-db66-4c93-a358-a562c069b1f1

Published study links

• Paper:https://www.sciencedirect.com/science/article/pii/S0092867421014811?via%3Dihub#sec5

• Accessioned data:

  • GSE192740
  • GSE192742
  • GSE192741

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

[ESapenaVentura]

Please note this dataset contains other organisms' data (e.g. Zebrafish)

Please don't include these when wrangling

[ipediez]

Needs term request:

• [ ] Liver abscess, as child of abscess [MONDO:0005227]
• [ ] Intersphincteric abscess, as child of abscess [MONDO:0005227]

At the earliest, with the HCAO release scheduled for April 20th. In the meantime, if you need to annotate for a release before that date, you could use Abscess (MONDO:0005227) and combine that with the anatomy terms UBERON:0002107 'liver' and UBERON:0004916 'anal sphincter', respectively.

For now we are going to release the dataset with the abscess term for both, and in the future we will update the samples H10 (intersphinteric abscess) and H33 (liver abscess)

[ipediez]

Patients have several diseases, for the specimen level I have only kept those diseases that might affect directly the liver, such as diabetes or morbid obesity

[gabsie]

blocked because of the schema update

[idazucchi]

Hi! I'm done with the secondary review. You were right about 666 being the appropriate number for this dataset: congratulations on wrangling this!

I've made some notes, most of the dataset is all right, there are a few instances where you swapped the protocols by mistake and some other notes are about information that could be added either to the donors or to the protocols. Under Specimen from donor there are a lot of notes because I would model the specimens differently and I tried to write it down in a way that would be easy to read. I'm sorry it's so long but I promise it's not bad. If there is anything that you want to discuss let me know!

Project info

• some grants are missing

Donor

• The weight and height are missing for all humans but they all have the BMI: is this to reduce the identifying information for the donors?
• can we record in the disease field the steatosis>10% (I'm not sure we have an ontology term), as well as the fibrosis status (we have an ontology term for it)? Mice
• the sex is missing for all mice, you can put unknown
• the developmental stage for all mice is adult
• mice age is either 36+5 or 24+5 weeks
• all the mice dead are dead at collection
• for visium mice: cause of death euthanised with co2
• western diet mice:
  • sex is male according to the paper
  • I would change the disease to non-alcoholic fatty liver disease MONDO:0013209 and non-alcoholic steatohepatitis MONDO:0007027 to reflect the effects of the diet
• pbs mice:
  • they are 29 weeks old according to the paper
  • you could add to the description for the time course or maybe ask the authors? It’s not really clear what they did, in the paper they mention a time frame of 2h and in geo is 24h

    Mice were treated with 3.5 mg/kg PBS and 2 h later, livers were harvested without the capsule

Collection protocol

• you could use the term biopsy for liverBiopsy

Specimen from donor organism

mice I would model the specimens differently, most of the notes are about that

• for visium mice:
  • healthyMouseLiver_5 should have 4 specimens, then corresponding to 1 imaged specimen each
• healthy mice 10x: should be a total of 12 specimens
  • 2 more specimens for healthyMouseLiver_1
  • 1 more specimen for healthyMouseLiver_2
  • 1 more specimen for healthyMouseLiver_3
  • 3 more specimens for healthyMouseLiver_4
  • 2 more specimens for perfusedHealthyMouseLiver_1
  • 2 more specimens for perfusedHealthyMouseLiver_2
  • 1 more specimen for perfusedHealthyMouseLiver_3
  • healthyMouse_6 and healthyMouse_8 are donors for ex vivo digestion, so they shouldn’t be perfused mice, this carries over to the cell suspension. Their collection protocol should be liverExtraction
• standard diet mice:
  • sdMouseLiver_2 should have liverPerfusionExtraction for collection protocol
  • I would add 2 more specimens for sdMouseLiver_1
• western diet mice:
  • I would change the disease to non-alcoholic fatty liver disease MONDO:0013209 and non-alcoholic steatohepatitis MONDO:0007027

• pbs mice:

  • they extracted just the liver, without the capsule. I don't think we have an ontology term for that, maybe you could add it to the description
  • pbsMouseLiver_2 should have liverPerfusionExtraction for collection protocol

  humans

  • can we record in the disease field the steatosis>10% (I'm not sure we have an ontology term), as well as the fibrosis status (we have an ontology term for it)?
• how many specimen come from H02? it's not clear from the information in GEO
• I think humanLiver_H14 has known disease Hypercholesterolemia not hyperlipidemia
• should humanLiver_H18, humanLiver_H07, humanLiver_H21 have ‘adenoma’ in the disease field? The paper says explicitly that liver biopsies were isolated from healthy adjacent tissue removed during liver resection due to colorectal cancer metastasis, but there is no information directly for liver adenoma
• for donors H37 and H38 do we assume that one sample was used both for visium and 10x? I would model it with separate specimens

Dissociation protocol

• you could add the publication doi for nuclei dissociation
• I would add to description of enzymaticDissociationInVivo the perfusion step, because it’s already an enzymatic digestion
• you could add a mechanical dissociation protocol to be applied to both in and ex vivo digestion, to indicate that the liver was cut up

Enrichment protocol

• I would delete FACS_nuclei and fill in the cell viability instead because it’s not selecting a certain type of cell? Alternatively add the positive sign to DAPI

Cell suspension

• INSDC sample accessions are missing
• for cells suspensions that were FACS-enriched the cell viability was assessed with Fixable viability dies mice
• healthy mice 10x:
  • the specimen for GSM5764250 should be healthyMouseLiverCapsule_1
  • the digestion protocol should be enzymaticDissociationExVivo for GSM5764270, GSM5764271, GSM5764274, GSM5764275, GSM5764278, GSM5764279, GSM5764285
• standard diet mice:
  • the digestion protocol should be enzymaticDissociationInVivo for GSM5764260, GSM5764261
  • the size enrichment is missing for GSM5764254, GSM5764255, GSM5764256, GSM5764257, GSM5764260, GSM5764261, GSM5764262, GSM5764263, GSM5764288, GSM5764290, GSM5764291, GSM5764294, GSM5764296
• western diet mice:
  • the size enrichment is missing for GSM5764289, GSM5764292, GSM5764293, GSM5764295, GSM5764297
• pbs mice:
  • the digestion protocol should be enzymaticDissociationInVivo for GSM5764264, GSM5764265, GSM5764266, GSM5764267
  • the size enrichment is missing for GSM5764258, GSM5764259, GSM5764264, GSM5764265, GSM5764266, GSM5764267 humans
• GSM5764394, GSM5764395, GSM5764396, GSM5764397, GSM5764398, GSM5764399, GSM5764400, GSM5764401 should have mechanicalDissociationNuclei for dissociation protocol
• the size enrichment is missing for GSM5764338, GSM5764341, GSM5764344, GSM5764346, GSM5764349, GSM5764351, GSM5764354, GSM5764356, GSM5764358, GSM5764361, GSM5764363, GSM5764366, GSM5764368, GSM5764370, GSM5764373, GSM5764375, GSM5764377, GSM5764380, GSM5764383, GSM5764384, GSM5764385, GSM5764386, GSM5764387, GSM5764388, GSM5764389, GSM5764390, GSM5764391, GSM5764392, GSM5764393

Imaging preparation protocol

• consider splitting the imaging protocols to indicate clearly which slides were marked with antibodies (just GSM5764423)

Imaged specimen

• GSM5764414, GSM5764415, GSM5764416, GSM5764417 should have different specimens as input, all from the same donor
• you could add the INSDC sample accessions

Imaging protocol

• the imaging indications hopefully apply to both Abs+/- slides

Image file

• all right!! I'm sorry this is so long :(

Library preparation

• visium
  • consider adding an additional visium protocol for the slides that were not Abs-marked
  • you can fill in the kit manufacturer and leave just the kit name in the retail name section
• cite-seq: you can add the description of the blocking abs

  When CITE-seq was to be performed, cells were then stained with 2.4G2 antibody to block Fc receptors and CITE-seq antibodies for 20mins at 4C, before being washed in excess PBS with 2% FCS and 2mM EDTA. Antibody details are included in the key resources table

Sequence file

• these cell suspensions GSM5764310, GSM5764312, GSM5764314, GSM5764315, GSM5764316, GSM5764323, GSM5764327 should have lp_nucleiv3 as library preparation

Analysis protocol

• the ontology term is missing
• computational methods are missing
• derivation process is missing
• the description doesn’t include filtering but the files are all filtered_features_bc, should this be a processed matrix protocol?

Analysis file

• the imaged specimen input for GSM5764423_counts_ADT_BH001.csv.gz should be GSM5764423
• for the analysis files from Cite-seq you list both cite-seq and 10x library prep: isn’t cite-seq enough?
• the cell count matrices are missing for all the human 10x CS, did you exclude them for a particular reason?

[ofanobilbao]

@ipediez and @idazucchi to have a meeting to go over some of the modelling. Irene has applied most of the changes.

[ipediez]

Waiting on contributor communications until Wednesday, as I contacted them two weeks ago for another issue and they haven't responded yet.

If they do not respond, the dataset is ready to go.

[idazucchi]

Will need a very small update, the cite-seq libraries have a 10x term, it should be EFO:0009294

[idazucchi]

Karoly changed the submission's status to Exported I filled out the import form for release 18 but Jeff Korte kindly suggested to move it to release 17

[Wkt8]

Reopening as this is the final 'double-pipes' issue which is as a result to project.contributors.laboratory not appearing in the UI.

Alexie and I edited the project in the API and it is ready to be exported + import form to be filled. Currently blocked by ontology release.

Tagging @gabsie for tracking.