search

broadinstitute / Tangram

Spatial alignment of single cell transcriptomic data.
BSD 3-Clause "New" or "Revised" License
246 stars 50 forks source link

Follow up on Spot-Level Deconvolution #36

Closed aseyedia closed 1 month ago

aseyedia commented 3 years ago

Hello,

My colleagues and I have been able to use Tangram to successfully map genes onto our Visium spatial dataset thanks to your incredibly helpful tutorial. However, we would like to implement deconvolution (i.e. assigning cell types to each spot of the Visium slide) as part of our Tangram pipeline in order to generate images similar to the one included in the manuscript and attached below for reference but we are unsure how to go about it. I was not able to find functions related to deconvolution in the Git repository with the exception of df_to_cell_types in utils.py:

https://github.com/broadinstitute/Tangram/blob/eb867f54c6e3a20872dbfc70bb7496f45d8a1dbe/tangram/utils.py#L376-L402

which is commented out. Is deconvolution forthcoming or am I overlooking something?

image

Like I mentioned before, we have otherwise been able to implement some of Tangram’s functionality already and we appreciate the help you’ve given us in the past.

Thank you, Arta Seyedian

lewlin commented 3 years ago

Woohoo - we love you enjoy the method.

We have not implemented yet deconvolution in "an easy" fashion BUT, the Squidpy team wrote a beautiful tutorial which I recommend

https://squidpy.readthedocs.io/en/stable/external_tutorials/tutorial_tangram.html

Let me know if this helps (we are working on a simpler version but it will take some time)

lewlin commented 2 years ago

Hi Enan,

I was responding on GitHub but I cannot see the message anymore: did you delete it?

For the histogram of cell types: if you do deconvolutions (ie you know how many cells you have in each spots), then computing an histogram of frequencies is very simple. Did you mean you needed help on the deconvolution instead?

Ciao

On Mon, Jan 17, 2022 at 8:06 AM Enan Ajmain @.***> wrote:

@lewlin https://github.com/lewlin, in that tutorial for deconvolution, I cannot find the steps to get celltype proportions. Could you help me on that?

— Reply to this email directly, view it on GitHub https://github.com/broadinstitute/Tangram/issues/36#issuecomment-1014694059, or unsubscribe https://github.com/notifications/unsubscribe-auth/ADYW7KHZNDPFOZTGM536RN3UWQ46PANCNFSM477FGD6Q . Triage notifications on the go with GitHub Mobile for iOS https://apps.apple.com/app/apple-store/id1477376905?ct=notification-email&mt=8&pt=524675 or Android https://play.google.com/store/apps/details?id=com.github.android&referrer=utm_campaign%3Dnotification-email%26utm_medium%3Demail%26utm_source%3Dgithub.

You are receiving this because you were mentioned.Message ID: @.***>