search

BiomedicalMachineLearning / stLearn

A novel machine learning pipeline to analyse spatial transcriptomics data
Other
201 stars 26 forks source link

Running time of PTST analysis for Xenium data #306

Open Ferris01 opened 4 months ago

Ferris01 commented 4 months ago

Hi developer team,

Thank you for developing this very useful tool. When I run st.spatial.trajectory.pseudotime function with my own Xenium data, which is clustered and annotated manually, it was super slow and it didn't end even after more than 10 hrs. Can you help to check whether it is the problem of the code or the size of the dataset. This dataset consists of 425 genes and 579517 cells. My codes are shown as below. Thank you for your help.

adata.uns['iroot']=st.spatial.trajectory.set_root( adata, use_label="cell_type_lvl5", cluster='B_C1_LRMP', use_raw=True ) st.spatial.trajectory.pseudotime(adata,eps=50,use_rep="X_pca",use_label="cell_type_lvl5")

Fabiulous93 commented 13 hours ago

I would be grateful if there have been any developments on this matter. I have encountered the same issue and would appreciate any solutions you may have found.