search

theislab / scvelo

RNA Velocity generalized through dynamical modeling
https://scvelo.org
BSD 3-Clause "New" or "Revised" License
408 stars 103 forks source link

velocity_pseudotime: from scv.tl.latent_time vs scv.tl.velocity_pseudotime #3588 #1168

Open savagyan00 opened 9 months ago

savagyan00 commented 9 months ago

I am running the basic pipeline (scvelo 0.3.1), starting from an AnnData object:

Preprocessing scv.pp.filter_and_normalize( adata_maca_10x, min_shared_counts=20, n_top_genes=2000, subset_highly_variable=False )

sc.tl.pca(adata_maca_10x) sc.pp.neighbors(adata_maca_10x, n_pcs=30, n_neighbors=30, random_state=0) scv.pp.moments(adata_maca_10x, n_pcs=None, n_neighbors=None) run velocity scv.tl.recover_dynamics(adata_maca_smartseq, n_jobs=8) scv.tl.velocity(adata_maca_10x, mode="dynamical") after this step I take one two ways

run latent_time : this generates a column "velocity pseudotime" in obs scv.tl.latent_time(adata_maca_10x)

(when I run velocity_pseudotime after this step without velocity_graph, also different values are generated)

run velocity pseudotime: this also generates a column "velocity pseudotime" in obs scv.tl.velocity_graph(adata_maca_10x) scv.tl.velocity_pseudotime(adata_maca_10x) But the values from these two are completely different, which I am unsure how to interpret. Can you please explain or point to documentation where I can find descriptions for these two value outputs?

Thanks!

sbombin commented 4 months ago

I have the same issue. Did you find the answer?