After training error: shapes not aligned; None of the alternative feature names were found in the dataset.

[ivanovaos]

Hi,

I tried to run the tool on a couple of more datasets, this is from tabula sapiens 2.0

copyvae Liver_TSP1_30_version2d_10X_smartseq_scvi_Nov122024.h5ad data/Macosko_cell_cycle_genes.txt -mc 6

The error message appears only at the end.

�[A
100%|██████████| 200/200 [24:06<00:00,  7.23s/epoch, loss=316]
Traceback (most recent call last):                            
  File "copyvae_cnv221/bin/copyvae", line 33, in <module>
    sys.exit(load_entry_point('copyVAE', 'console_scripts', 'copyvae')())
  File "temp_copyvae/copyVAE-main/copyvae/pipeline.py", line 305, in main
    run_pipeline(file, cc_genes, bin_size, max_cp, intermediate_dim, latent_dim, batch_size, epochs, number_of_clones)
  File "temp_copyvae/copyVAE-main/copyvae/pipeline.py", line 198, in run_pipeline
    cluster_auto_corr, ncell_index = find_normal_cluster(x_bin, pred_label)
  File "temp_copyvae/copyVAE-main/copyvae/clustering.py", line 89, in find_normal_cluster
    cluster_auto_corr[i] = auto_corr(cluster_x - cluster_mean)
  File "temp_copyvae/copyVAE-main/copyvae/clustering.py", line 61, in auto_corr
    res += np.sum(cluster_data[i,:] * cluster_data[j,:])
  File "copyvae_cnv221/lib/python3.9/site-packages/numpy/matrixlib/defmatrix.py", line 221, in __mul__
    return N.dot(self, asmatrix(other))
ValueError: shapes (1,1370) and (1,1370) not aligned: 1370 (dim 1) != 1 (dim 0)

And this is on the https://www.nature.com/articles/s41591-024-03376-x

copyvae final_ec_oct2024_cxg.h5ad data/Macosko_cell_cycle_genes.txt -mc 6

Traceback (most recent call last):
  File "/copyvae_cnv221/bin/copyvae", line 33, in <module>
    sys.exit(load_entry_point('copyVAE', 'console_scripts', 'copyvae')())
  File "temp_copyvae/copyVAE-main/copyvae/pipeline.py", line 305, in main
    run_pipeline(file, cc_genes, bin_size, max_cp, intermediate_dim, latent_dim, batch_size, epochs, number_of_clones)
  File "temp_copyvae/copyVAE-main/copyvae/pipeline.py", line 161, in run_pipeline
    adata = filter_and_normalize_data(umi_counts,
  File "temp_copyvae/copyVAE-main/copyvae/preprocess.py", line 60, in filter_and_normalize_data
    raise ValueError("None of the alternative feature names were found in the dataset.")
ValueError: None of the alternative feature names were found in the dataset.

[Jessica1525809]

@kurtsemih @mandichen I encountered the same problem. Have you solved it?