error drawing the plot for P1, P12, P100, mouse, scRNA-seq, utricular hair cells (Kelley)

[beamilon]

@adkinsrs When doing a gene search with single gene view, there is a message "There was an error drawing the plot." for P1, P12, P100, mouse, scRNA-seq, utricular hair cells (Kelley). Vestibular profile. Looking if other datasets have the same problem. I started a list below but realized that it involves all the SVGs.

[beamilon]

Additional datasets (some are in several profiles but I listed them only once): I stopped adding more because it seems that all the SVGs are broken.

• scRNA-Seq of FACS-purified adult mouse Lfng-EGFP cochlear supporting cells(log-normalized, filtered, Dr.Michael Hoa) in "Adult"
• E6.5, chicken, RNA-seq, proximal, middle, distal sections of basilar papilla (Corwin) in "Chick"
• E10.5, mouse, microarray, otic vesicle, periotic tissue, and non-otic tissue (Heller) and E12.5-P30, mouse, scRNA-seq, endolymphatic sac (Griffith) in "Development"
• P0, mouse, RNA-seq, cochlea, hair cells compared with rest of cochlear duct (Groves) in "Hair cell"
• P0, mouse, RNA-seq, hair cells vs epithelial non-hair cells (Hertzano) and P2, mouse, sc-qRT-PCR, cochlea (Heller) and E16.5, P0, mouse, RNA-seq, cochlear and vestibular sensory epithelium (Avraham) in Hearing (default)
• P30, mouse, sc/snRNA-seq, comparison of approaches (Hoa) in "Lateral wall"
• The Adult Cochlea Response to PTS-Inducing Noise - Summary View (Log2) and The Adult Cochlea Response to PTS-Inducing Noise – Summary View (24h time point, vector graphic) in "NIHL, cell type-specific transcriptomics, mouse cochlea (Hertzano 2021)
• P0, mouse, RNA-seq, epithelial vs neuronal, mesenchymal and vascular endothelial (Hertzano) in "Other"

[adkinsrs]

Can you check some of these now to see if they work? I noticed that some of the more recently uploaded svgs were owned by the root user rather than @jorvis . I changed the Unix user of these SVGs to @jorvis and I now can see the image from the first comment

[beamilon]

It seems to be working now but I didn't check them all. It was for all the SVGs, not only the newer ones.

[beamilon]

4dpf, zebrafish, microarray, hair cell, mantle cell and skin (Hudspeth) in zebrafish is still broken. Actually, the other ones disappeared again and I had to refresh again to get them to reappear

[adkinsrs]

Interesting, since I can see that one without issue. Can you clear your cache to see if that resolves the issue?

[adkinsrs]

Or @beamilon is it happening with a particular gene that you are searching for?

[beamilon]

It's fine now. For some reason all the SVGs disappeared again so I had to refresh one more time for them to reappear.

[beamilon]

They are gone again!! I need to refresh every few minutes or they disappear!

[beamilon]

Yep. Gone again. Reappear only if I refresh again.

[adkinsrs]

Going to rope in @jorvis and see if he has any ideas. I wonder if there is a resource limit on the server being hit. I have a "try/catch" block that is executed in js/classes/display.js/draw() that either displays an error message from the Python API call or a generic "There was an error drawing the plot" message.

@beamilon can you go into the console again (Right-click -> Inspect -> Console tab) and print out any potential messages there. I am wondering if the JS AbortController() event is firing but acting oddly (missing variable or syntax or something).

[beamilon]

The only things I see seem to be related to Epiviz, even if I am in a page without Epiviz.

[jorvis]

@beamilon I need a clarification here. When you say 'disappear' do you mean that you have the page loaded, are looking at things, and the images suddenly disappear or do you mean that when you're searching/clicking different genes or changing profiles sometimes they load and sometimes they don't?

[beamilon]

What I mean is that the message "There was an error drawing the plot." returns instead of the plot.

[beamilon]

So, I played with it more. I was trying to replicate the "There was an error drawing the plot." after changing profile several times and every thing worked great. Then I realized that if I used a profile with an Epiviz dataset, and got the error message below, after that, all the SVGs returns the "There was an error drawing the plot." message instead of loading the image. Here is the URL: https://umgear.org/index.html?multigene_plots=0&layout_id=f64f9c22&gene_symbol_exact_match=1&gene_symbol=sox2,pou4f3,gfi1

[beamilon]

The URL works perfectly with SVGs though, unless you first try a profile that triggers the Epiviz error message and then try to get the SVGs.

[jorvis]

OK, progress. And which epiviz-containing profile are you swapping to?

[beamilon]

Hair cell and supporting cell transcriptomics and epigenomics, Mouse (Segil 2021) or Hair cell reprograming - transcriptomics and epigenetics (Segil).

[beamilon]

Is there any updates on that issue?

[jorvis]

Not yet, I'll try to return to it soon.