Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

How to set the simularity threashold for treeplot #4

Open
cchd0001 opened this issue Jan 30, 2024 · 2 comments
Open

How to set the simularity threashold for treeplot #4

cchd0001 opened this issue Jan 30, 2024 · 2 comments

Comments

@cchd0001
Copy link

Hi dear developer,

First of all, thanks for creating this valuable tool, really appreciate it a lot.

Here is my issue: the similarity of the two classes is not too low (0.464371) but no link is displayed in the treeplot.
How can I reduce this similarity threshold to increase the links of treeplot ?

Best wishes,
Lidong Guo

@ChuanXu1
Copy link
Collaborator

ChuanXu1 commented Feb 12, 2024

@cchd0001, if two classes A and B are not matched, it is likely that one of them (for example, A) aligns with another cell type C which has a greater similarity compared to that between A and B. Otherwise, it is probable that these two cell types are not quite similar in the latent space (replacing the latent space with a better one such as that after batch correction will also yield optimal results sometimes). You can also adjust the parameter minimum_unique_percents in cellhint.harmonize to examine the outputs, as this will impact the threshold in cell type alignment.

@cchd0001
Copy link
Author

cchd0001 commented Feb 18, 2024

Thanks.

In my case, A (from dataset_01) and B (from dataset_02) are matched if only this two dataset (dataset_01 and dataset_02) harmonize. Then I add dataset_03, A matched celltype C in dataset_03 but didn't match B any more.

BTW, the final order is dataset_02, dataset_01, dataset_03.

I will try this minimum_unique_percents.

Thanks again

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants