Change plot title from Disease vs GTEX to the particular disease referenced by the EFO_ID

[dunnpa]

The plot has title "Disease vs GTEX tissue bulk gene expression"

The API call has the efo_id. Can that can be used to get the corresponding disease name and enter that into the plot title?

[jharenza]

@dunnpa thanks for noticing this! I would suggest we use the cancer group instead of disease that is the result of the EFO ID, as these are a bit off from what we'd like to see (eg - we have had a cancer group we had to lump into an EFO id named for "CNS neoplasm" or multiple into one EFO id since all weren't available, and that's not useful, but we had to do it that way to get some EFO id)

[dunnpa]

@jharenza, I assume there is a mapping of efo_id to cancer group to do as you suggest? I was going by the available API parameters, but your suggest offers a finer resolution name for the plot.

[logstar]

The plot has title "Disease vs GTEX tissue bulk gene expression"

The API call has the efo_id. Can that can be used to get the corresponding disease name and enter that into the plot title?

@dunnpa Thank you for the question. The plot title can surely be changed. Certain EFO IDs are mapped to multiple diseases, e.g. https://openpedcan-api-qa.d3b.io/tpm/gene-disease-gtex/plot?ensemblId=ENSG00000157764&efoId=EFO_0000514. Would collapsing them by " and " work? For example, "Germinoma and Teratoma vs GTEX tissue bulk gene expression".

(EDIT) *diseases should be cancer groups. I assume one EFO ID maps to one disease on PedOT.

@dunnpa thanks for noticing this! I would suggest we use the cancer group instead of disease that is the result of the EFO ID, as these are a bit off from what we'd like to see (eg - we have had a cancer group we had to lump into an EFO id named for "CNS neoplasm" or multiple into one EFO id since all weren't available, and that's not useful, but we had to do it that way to get some EFO id)

@jharenza Would "OpenPedCan cancer group vs GTEX tissue bulk gene expression" be an informative plot title?

[logstar]

@jharenza, I assume there is a mapping of efo_id to cancer group to do as you suggest? I was going by the available API parameters, but your suggest offers a finer resolution name for the plot.

@dunnpa I think @jharenza was suggesting to distinguish PedOT "disease" from OpenPedCan "cancer group".

I also think maybe we should use "OpenPedCan cancer group" to inform PedOT users that the boxes may not represent the PedOT "disease" exactly.

[jharenza]

I was suggesting in this plot, for instance, using "Neuroblastoma vs GTEX..." or whatever is in cancer group rather than EFO

[logstar]

I was suggesting in this plot, for instance, using "Neuroblastoma vs GTEX..." or whatever is in cancer group rather than EFO

@jharenza Thank you. The API currently only supports query by EFO ID. The API could support query by cancer group, but it would need to be coordinated with the PedOT front-end, where the API is directly queried. The PedOT front-end only has EFO ID information but does not have cancer group information, e.g. https://ppdc-otp-stage.bento-tools.org/evidence/ENSG00000273032/EFO_0000616, so the API is currently designed to only support query by EFO ID in https://nih.box.com/s/5cq2jwi6bhg0mgnowad3e6e4i60hwbnr.

@jharenza @dunnpa Do the following example titles look good?

• "OpenPedCan Germinoma and Teratoma vs GTEX tissue bulk gene expression" for EFO_0000514. Example URL with current title, https://openpedcan-api-qa.d3b.io/tpm/gene-disease-gtex/plot?ensemblId=ENSG00000157764&efoId=EFO_0000514
• "OpenPedCan Neuroblastoma vs GTEX tissue bulk gene expression" for EFO_0000621. Example URL, https://openpedcan-api-qa.d3b.io/tpm/gene-disease-gtex/plot?ensemblId=ENSG00000157764&efoId=EFO_0000621.

[jharenza]

Ahh I see. In that case, I suppose those titles are fine, or we can even say "tumor" vs GTEX (which goes back to the original) so we don't have to add the diseases since they are on the x-axis. I worry the titles will become very bulky and as long as we are separating cancer groups within the plots, this should be ok. A next step would be subgroups within a cancer - say for NBL, we have MYCN amp and non amp and/or high risk and low risk (or even high risk MYCN amp, etc). They'll all be the same EFO for some time, so I would suggest then just bring general here and not adding the name of the disease.

[logstar]

Ahh I see. In that case, I suppose those titles are fine, or we can even say "tumor" vs GTEX (which goes back to the original) so we don't have to add the diseases since they are on the x-axis. I worry the titles will become very bulky and as long as we are separating cancer groups within the plots, this should be ok. A next step would be subgroups within a cancer - say for NBL, we have MYCN amp and non amp and/or high risk and low risk (or even high risk MYCN amp, etc). They'll all be the same EFO for some time, so I would suggest then just bring general here and not adding the name of the disease.

Thank you very much for the details. I agree. I always forgot about the subgoups, which are very relevant here.

@jharenza @dunnpa Does the following title look good, "Tumor vs GTEx tissue bulk gene expression"?

[logstar]

@jharenza @dunnpa - By the way, does the following title look good for pan-cancer plots, "Tumor tissue bulk gene expression"?

I noticed that this issue is only about tumor vs gtex plots, but I think the pan-cancer plots are also related.

[jharenza]

We might actually write "primary tumor" to designate phase of therapy - the rest sounds fine

[logstar]

We might actually write "primary tumor" to designate phase of therapy - the rest sounds fine

Thank you for the suggestion. I agree primary tumor is more informative. Following are the updated titles:

• "Primary tumor vs GTEx tissue bulk gene expression" for tumor vs gtex boxplot.
• "Primary tumor tissue bulk gene expression" for pan-cancer boxplot.

I will implement these two titles in the next PR, but they could still be changed at a later point.

[taylordm]

Let’s double check with the clinicians on the titles.

…

On Sep 1, 2021, at 6:21 PM, Yuanchao Zhang ***@***.***> wrote:  @jharenza @dunnpa - By the way, does the following title look good for pan-cancer plots, "Tumor tissue bulk gene expression"? I noticed that this issue is only about tumor vs gtex plots, but I think the pan-cancer plots are also related. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub, or unsubscribe. Triage notifications on the go with GitHub Mobile for iOS or Android.

[chinwallaa]

@logstar will follow up with FNL regarding next round of API updates on FNL side.

[chinwallaa]

discussion - when we update API, will follow up with FNL regarding plot title changes