why my adata.obs including Clusters, _X and _Y ?

Hi, Thank you for creating scvelo. That is a very great tool.

I am a novice for single cell analysis. So my problem may be ridiculous. But it really perplexed me.

I ran Cellranger count to get matrix and I used velocyto to align. And finally, I got a .loom file. I checked this file and found out there are columns such Clusters, _X ,_Y in my adata.obs.

I feel puzzled. I didn’t do cluster and why there are Clusters, _X, _Y ?

My code is displayed as following:

houruiyan@GPU:~/10genomic$ nohup cellranger count --id=run_count_1kpbmcs --fastqs=./pbmc_1k_v3_fastqs --sample=pbmc_1k_v3 --transcriptome=./refdata-cellranger-GRCh38-3.0.0 & (velocyto) houruiyan@GPU:~/10genomic$ nohup velocyto run10x ~/10genomic/run_count_1kpbmcs/ /home/houruiyan/10genomic/Homo_sapiens.GRCh38.93.gtf &

[houruiyan]

Thank you very much !!! Thank you for your patient explanation!

[VolkerBergen]

From the deterministic/steady-state model, you only obtain one value of gamma denoted as ‘velocity_gamma’, which is the relative steady-state ratio, not the degradation rate.

The dynamical model provides many estimates all starting with ‘fit_’, including the rates of transcription (‘fit_alpha’), splicing (‘fit_beta’) and degradation (‘fit_gamma’).

In the latest source version, you can fetch all anndata keys that start with ‘fit_’ by path search via scv.get_df(adata, 'fit_*').

Per default, the dynamical only fits phase trajectories for those genes, that can be reliably fitted. You can force it to fit all genes by setting var_names='all' in tl.recover_dynamics.