Downstream scRNA UMAP analysis

[ninashenker]

Hi,

I'm trying to recreate your mouse hypothalamus UMAP in the paper (Figure 5b). Here you show UMAP layout of single-cell profiles generated from bulk tissue by Bulk2Space.

I have been unsuccessful in recreating the UMAP. I've tried a few thing such as the following:

adata= sc.AnnData(generate_sc_data.T)
generate_sc_meta.set_index('Cell', inplace=True)
adata.obs['Cell_type'] = generate_sc_meta['Cell_type']
sc.pp.neighbors(adata)
sc.tl.umap(adata)
sc.pl.umap(adata, color='Cell_type')

The UMAP shows no clustering or separation of cell types.
Could you please provide the code for generating a UMAP after deconvolution? Or assist me in how to correctly get the data into adata format?

I would really appreciate any help!

Best,
Nina