Problem in runing RSS in 02_preprocessing_and_integration.ipynb

[XuYingJie-z]

Hello HNOCA Team,

I’m really excited that your group has integrated these organoid data. I would really like to reproduce this paper. I am having trouble running the code in 02 preprocessing and integration.ipynb.

Problem:
Where can I find HumanFetalBrainPool_cluster_expr_highvar.tsv？

# run RSS to Braun et al. 
ref = pd.read_csv("/storage/data/organoid_atlas_adatas/HumanFetalBrainPool_cluster_expr_highvar.tsv", sep="\t")
ref = ref.loc[np.isin(ref.index, adata.var_names),:]

ref_ranked = rankMatrix_nonzero(ref)
ranked_X = rankMatrix_nonzero(adata[:,ref.index].X.T)

cor2ref_spearman = corSparse(ranked_X, ref_ranked)
cor2ref_spearman_scaled = scale(cor2ref_spearman, axis = 1)
cor2ref_spearman_scaled[np.isnan(cor2ref_spearman_scaled)] = 0

pca_rss = PCA(n_components=20).fit_transform(cor2ref_spearman_scaled)

Thank you!

[zhisonghe]

Hi,

Sorry for the omission. We have now included the file in the repository (supplemental_files/HumanFetalBrainPool_cluster_expr_highvar.tsv). It was the average expression matrix of their highly variable genes across all clusters reported by Braun et al. 2023 (the preprint version in biorxiv).

All the best,
Zhisong