Problem with merge function

[ljouneau]

Hello,

After having loaded 2 velocyto loom files, I tried to merge them but the result shows very few (22) observations

import scvelo as scv
adata_t0 = scv.read("../Analyses_velocyto/icare5_t0h/velocyto/icare5_t0h.loom",cache=True)
adata_t4 = scv.read("../Analyses_velocyto/icare5_t4h/velocyto/icare5_t4h.loom",cache=True)
adata=scv.utils.merge(adata_t0,adata_t4)
adata_t0
AnnData object with n_obs × n_vars = 8288 × 60672
var: 'Accession', 'Chromosome', 'End', 'Start', 'Strand'
layers: 'ambiguous', 'matrix', 'spliced', 'unspliced'
adata_t4
AnnData object with n_obs × n_vars = 7841 × 60672
obs: 'initial_size_spliced', 'initial_size_unspliced', 'initial_size'
var: 'Accession', 'Chromosome', 'End', 'Start', 'Strand'
layers: 'ambiguous', 'matrix', 'spliced', 'unspliced'
adata
AnnData object with n_obs × n_vars = 22 × 60672
obs: 'initial_size_spliced', 'initial_size_unspliced', 'initial_size'
var: 'Accession', 'Chromosome', 'End', 'Start', 'Strand'
layers: 'ambiguous', 'matrix', 'spliced', 'unspliced'

Thank you for your answers

[WeilerP]

Not sure what the question is but I presume it's why there are only 22 observations left after merging. Looks like your two AnnData objects only have 22 observations in common. You can check the number of common observations yourself by

import pandas as pd
import scvelo as scv

adata_t0 = scv.read("../Analyses_velocyto/icare5_t0h/velocyto/icare5_t0h.loom",cache=True)
adata_t4 = scv.read("../Analyses_velocyto/icare5_t4h/velocyto/icare5_t4h.loom",cache=True)

pd.unique(adata_t0.obs_names.intersection(adata_t4.obs_names)).size