search

Cellular-Semantics / whole_mouse_brain_ontology

A data-driven cell-type ontology of the whole mouse brain.
Apache License 2.0
0 stars 0 forks source link

Add NS-Forest markers (once available) #21

Open dosumis opened 5 days ago

dosumis commented 5 days ago

@yunzhang813

The Whole Mouse Brain taxonomy (https://www.nature.com/articles/s41586-023-06812-z) datasets are ready to download. We split the taxonomy based on taxonomy classes (total 34) for convenience. You can find the download links below:

https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_01.h5ad https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_02.h5ad https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_03.h5ad ... https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_33.h5ad https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_34.h5ad

Do you think worth generating NS-Forest markers in the context of each class? These can be folded into our ontology product for BICAN as well as into the CL KG

Could also potentially provide confidence scores for combo.markers in these tables: https://static-content.springer.com/esm/art%3A10.1038%2Fs41586-023-06812-z/MediaObjects/41586_2023_6812_MOESM8_ESM.xlsx.

Again - from these files, the context would be per Class.

yunzhang813 commented 4 days ago

We have generated NS-Forest markers already for the 34 classes. The author marker table does not provide markers.combo for classes.

Actually, we have also generated NS-Forest markers for the 338 subclasses. We can do an evaluation for author’s subclass.markers.combo.

Please let me know your thoughts.

From: David Osumi-Sutherland @.> Date: Thursday, November 21, 2024 at 1:45 PM To: Cellular-Semantics/whole_mouse_brain_ontology @.> Cc: Zhang, Yun (Renee) @.>, Mention @.> Subject: [Cellular-Semantics/whole_mouse_brain_ontology] Add NS-Forest markers (once available) (Issue #21)

@yunzhang813https://github.com/yunzhang813

The Whole Mouse Brain taxonomy (https://www.nature.com/articles/s41586-023-06812-z) datasets are ready to download. We split the taxonomy based on taxonomy classes (total 34) for convenience. You can find the download links below:

https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_01.h5ad https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_02.h5ad https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_03.h5ad ... https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_33.h5ad https://cellular-semantics.cog.sanger.ac.uk/public/merged_CS20230722_CLAS_34.h5ad

Do you think worth generating NS-Forest markers in the context of each class?

Could also potentially provide confidence scores for combo.markers in these tables: https://static-content.springer.com/esm/art%3A10.1038%2Fs41586-023-06812-z/MediaObjects/41586_2023_6812_MOESM8_ESM.xlsx.

Again - from these files, the context would be per Class.

— Reply to this email directly, view it on GitHubhttps://github.com/Cellular-Semantics/whole_mouse_brain_ontology/issues/21, or unsubscribehttps://github.com/notifications/unsubscribe-auth/ADMLYMXCE4ZT46Y5L7PJMBL2BZHRJAVCNFSM6AAAAABSH45FZ2VHI2DSMVQWIX3LMV43ASLTON2WKOZSGY4DAOJSHEZTAMA. You are receiving this because you were mentioned.Message ID: @.***>

dosumis commented 4 days ago

That's great. Can you share NS-Forest maerk via CellMark? Would be great to get confidence scores too. We can incorporate all of this in the ontology pretty rapidly as we already have infrastructure in place.

yunzhang813 commented 3 days ago

Here are the NS-Forest markers for the 34 classes. WMBTaxonamy_global_class_results.csv In the table, NS-Forest markers are in the column "markers" and the confidence scores are in the column "f_score".

The author marker table does not provide markers.combo for classes.