The integrated showing error

[honghh2018]

Hi guy,
When i runing on below step,
adatas_cor = scanorama.correct_scanpy(adatas, return_dimred=True)
It was complain about error:
Error: No genes found in all datasets, exiting...

NameError Traceback (most recent call last)
/tmp/ipykernel_146881/3261772617.py in
----> 1 adatas_cor = scanorama.correct_scanpy(adatas, return_dimred=True)

/share/nas1/Data/software/mini_conda/Miniconda3/envs/Scanpy_package/lib/python3.8/site-packages/scanorama/scanorama.py in correct_scanpy(adatas, **kwargs)
219 """
220 if 'return_dimred' in kwargs and kwargs['return_dimred']:
--> 221 datasets_dimred, datasets, genes = correct(
222 [adata.X for adata in adatas],
223 [adata.var_names.values for adata in adatas],

/share/nas1/Data/software/mini_conda/Miniconda3/envs/Scanpy_package/lib/python3.8/site-packages/scanorama/scanorama.py in correct(datasets_full, genes_list, return_dimred, batch_size, verbose, ds_names, dimred, approx, sigma, alpha, knn, return_dense, hvg, union, seed)
90 datasets_full = check_datasets(datasets_full)
91
---> 92 datasets, genes = merge_datasets(datasets_full, genes_list,
93 ds_names=ds_names, union=union)
94 datasets_dimred, genes = process_data(datasets, genes, hvg=hvg,

/share/nas1/Data/software/mini_conda/Miniconda3/envs/Scanpy_package/lib/python3.8/site-packages/scanorama/scanorama.py in merge_datasets(datasets, genes, ds_names, verbose, union)
336 if len(keep_genes) == 0:
337 print('Error: No genes found in all datasets, exiting...')
--> 338 exit(1)
339 if verbose:
340 print('Found {} genes among all datasets'

NameError: name 'exit' is not defined

So how can i fix this issue?
Any advice would be appreciated

[brianhie]

Hi @honghh2018, could you make sure that the genes you are passing to scanorama are the same? By default, it will take the intersection of all gene sets.

[honghh2018]

So, how can i add intersection of the two samples genes in adatas just like i post above object.
Best,
hanhuihong

[brianhie]

You can pass in the gene sets as adatas. Just make sure that

len(set(adata1.var_names) & set(adata2.var_names)) > 0

[honghh2018]

You mean like below, using the intersection genes that within in two adata objects? the command lying below:
adatas_cor = scanorama.correct_scanpy(adatas, return_dimred=True, genes[set(adata1.var_names) & set(adata2.var_names)])

I would try and feedback soon.
Best

[honghh2018]

sorry for the late reply,
I found the something weird that vignette post in you homepage about data integrated by scanorama.
https://nbisweden.github.io/workshop-scRNAseq/labs/compiled/scanpy/scanpy_03_integration.html
It was too complicated to figure it out myself.

in my code, i want to integrated two single cell sample by scanorama algorithm was processing into below step:
STIM=sc.read_csv("./t2_STIM.xls").T
ctrl=sc.read_csv("./t2_ctrl.xls").T
STIM.var_names_make_unique()
ctrl.var_names_make_unique()
sc.pp.calculate_qc_metrics(STIM, inplace=True)
sc.pp.calculate_qc_metrics(ctrl, inplace=True)
for adata in [
STIM,
ctrl,
]:
sc.pp.normalize_total(adata, inplace=True)
sc.pp.log1p(adata)
sc.pp.highly_variable_genes(adata, flavor="seurat", n_top_genes=2000, inplace=True)

adatas = [STIM, ctrl]
adatas_cor = scanorama.correct_scanpy(adatas, return_dimred=True)

and then, I can real not how to step next when i process my datas come at this step:
adatas_cor = scanorama.correct_scanpy(adatas, return_dimred=True)
Would you please give me some tips to kick this issue off?
Best,
Hanhuihong

[honghh2018]

@brianhie
Had any advice about my issue?
Is that right about my code?
Best