IGS / gEAR

The gEAR Portal was created as a data archive and viewer for gene expression data including microarrays, bulk RNA-Seq, single-cell RNA-Seq and more.
https://umgear.org
GNU Affero General Public License v3.0
13 stars 4 forks source link

Related to #356 that was closed #362

Open beamilon opened 2 years ago

beamilon commented 2 years ago

@songeric1107 Display not working for Mouse, scRNAseq,postnatal mouse utricle,raw matrix (Jan) image

jorvis commented 2 years ago

Was this dataset changed on the back-end (maybe updated by Yang?) Possible fix here is to re-curate with current columns and set that one as the new primary curation.

beamilon commented 2 years ago

Yang did something to add real names for the clusters because they were just labelled S1, S2...but I don't know what was done for that.

songeric1107 commented 2 years ago

@jorvis , I think there might be a bug to displaying umap being generated using the sc workbench on the main page.

I did not change anything about this dataset. As you see from the title, this dataset is loaded with a raw matrix, the user needs to use sc workbench to analyze it before curating a plot.

I try to curate again, I am able to see the thumbnail plot, but not able to see it in the front page.

Screen Shot 2022-07-14 at 4 39 15 PM

Screen Shot 2022-07-14 at 4 37 41 PM

Screen Shot 2022-07-14 at 4 37 49 PM

songeric1107 commented 2 years ago

@beamilon , I did not change anything of this dataset, as you see from the title, it is loaded as raw matrix, not cell type information at all. The cell type information is added to another dataset in the same profile with "SWNE plot" in the title Screen Shot 2022-07-14 at 4 42 50 PM

beamilon commented 2 years ago

@songeric1107 The SWNE plot had the cell types already (it actually had like 30 clusters with names or so). The one that was missing the names of the cell types was Mouse, scRNAseq,postnatal mouse utricle,raw matrix (Jan), the one I mentioned in #356. It had only S1, S2...and so on but no corresponding names.

songeric1107 commented 2 years ago

@beamilon , as you see in the profile, I uploaded the datasets in two different ways:

  1. upload the 10x mtx matrix only (raw matrix), which needs to be analyzed using gEAR single-cell workbench. As you can see, the UMAP is created based on a saved analysis on 7-26. the cluster is only named in the numeric way(0-12), not with the cell type name, therefore, no cell type will display in the profile page. I don't know how to add the cell type info to the saved analysis. @jorvis , let me know if there is a trick to do so. Screen Shot 2022-07-14 at 4 54 26 PM

Screen Shot 2022-07-14 at 4 54 18 PM

  1. upload the processed results to gEAR to display the umap as the user created (SWNE plot)

If you need to add the celltype information to the saved analysis result, I may need help from Josh.

adkinsrs commented 2 years ago

Assigning myself, as I believe the first image may be a bug in light of the 2nd image using those adata.uns params. I think the saved analysis ID information is not being completely transferred to the would-be gene search display tSNE entry

adkinsrs commented 2 years ago

Also what profile is this (so I'm not looking through every profile)

songeric1107 commented 2 years ago

|profile || gEAR ID| scRNAseq_mouse_utricle(Jan) / Utricle SC / Vestibular|| e084843c-32b0-4551-7307-0942eaa45756

adkinsrs commented 2 years ago
Screen Shot 2022-07-15 at 9 42 09 AM

Good news, I got it working

Bad news, for some reason the panel header seems to stretch further vertically. Nothing to do with the plotting, but I think there is some bad CSS rules happening.

beamilon commented 2 years ago

@songeric1107 What Ronna wanted for #356 is to obtain the information so that the clusters are named instead of just numbered in Mouse, scRNAseq,postnatal mouse utricle,raw matrix (Jan) image

jorvis commented 2 years ago

@beamilon That seems like quite a different issue than this ticket. Is this already discussed elsewhere in another?

beamilon commented 2 years ago

356 is what Ronna wanted with Mouse, scRNAseq,postnatal mouse utricle,raw matrix (Jan). Then Mouse, scRNAseq,postnatal mouse utricle,raw matrix (Jan) display stopped working so I thought that it was because the dataset had been modified. There was some misunderstanding so both tickets became untangled.

songeric1107 commented 2 years ago

@beamilon @jorvis should we remove the raw matrix dataset from the public profile(vestibular) since the raw matrix display only aimed to provide the opportunity for users to run our sc-workbench. the "Mouse, scRNAseq,postnatal mouse utricle,SWNE plot (Jan)" on the 2nd row is the same dataset but labeled with the celltype information from the owner.

Screen Shot 2022-07-20 at 12 14 03 PM

songeric1107 commented 2 years ago

@jorvis Ronna agrees to remove the raw matrix display from the public "Vestibular" profile .

profile ID. https://umgear.org/p?l=c60474c6

The link for raw matrix display.

https://umgear.org/dataset_curator.html#/dataset/e084843c-32b0-4551-7307-0942eaa45756/displays