Bug report [BUG] Tutorial throws pyarrow.lib.ArrowTypeError: Expected bytes, got a 'int' object

[arne-cl]

Describe the bug
Cell [17] of the human cerebellum pycisTopic tutorial does not produce a TSS annotation BED file but throws a "pyarrow.lib.ArrowTypeError: Expected bytes, got a 'int' object".

To Reproduce
Install scenicplus via conda using the official instructions and follow the pycisTopic tutorial up to cell [17].

Error output

- Get TSS annotation from Ensembl BioMart with the following settings:
  - biomart_name: "hsapiens_gene_ensembl"
  - biomart_host: "http://www.ensembl.org/"
  - transcript_type: ['protein_coding']
  - use_cache: True
/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/pybiomart/dataset.py:269: DtypeWarning: Columns (0) have mixed types. Specify dtype option on import or set low_memory=False.
  result = pd.read_csv(StringIO(response.text), sep='\t')
Traceback (most recent call last):
  File "/home/arne/miniconda3/envs/scenicplus/bin/pycistopic", line 7, in <module>
    sys.exit(main())
             ^^^^^^
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/pycisTopic/cli/pycistopic.py", line 26, in main
    args.func(args)
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/pycisTopic/cli/subcommand/tss.py", line 459, in run_tss_get_tss_annotation
    get_tss_annotation_bed_file(
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/pycisTopic/cli/subcommand/tss.py", line 164, in get_tss_annotation_bed_file
    tss_annotation_bed_df_pl = ga.get_tss_annotation_from_ensembl(
                               ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/pycisTopic/gene_annotation.py", line 172, in get_tss_annotation_from_ensembl
    ensembl_tss_annotation_bed_df_pl = pl.from_pandas(ensembl_tss_annotation).select(
                                       ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/polars/convert.py", line 719, in from_pandas
    return pl.DataFrame._from_pandas(
           ^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/polars/dataframe/frame.py", line 621, in _from_pandas
    pandas_to_pydf(
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/polars/utils/_construction.py", line 1837, in pandas_to_pydf
    arrow_dict[str(col)] = _pandas_series_to_arrow(
                           ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/arne/miniconda3/envs/scenicplus/lib/python3.11/site-packages/polars/utils/_construction.py", line 665, in _pandas_series_to_arrow
    return pa.array(values, pa.large_utf8(), from_pandas=nan_to_null)
           ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "pyarrow/array.pxi", line 340, in pyarrow.lib.array
  File "pyarrow/array.pxi", line 86, in pyarrow.lib._ndarray_to_array
  File "pyarrow/error.pxi", line 91, in pyarrow.lib.check_status
pyarrow.lib.ArrowTypeError: Expected bytes, got a 'int' object

Expected behavior
I expected the TSS annotation BED file to be written to "outs/qc/tss.bed".

Version (please complete the following information):

• Python 3.11.8
• pycisTopic 2.0a0

Additional context
Add any other context about the problem here.

[arne-cl]

Note: #186 mentions a different error in the same cell. As a workaround, it was suggested to use the polars_1xx branch of pycisTopic.

Therefore, I set up a new conda env into which I only installed pycisTopic from the polars_1xx. Unfortunately, running the tutorial will still throw the same "pyarrow.lib.ArrowTypeError: Expected bytes, got a 'int' object" error.

[ghuls]

It is fixed in the pycistopic_v3 branch.

23638d8

[ckntav]

Hi,

I switched to the pycistopic_v3 branch, but I’m unable to run the tutorial because the pseudobulk_peak_calling module appears to be missing. This module is required for functions like export_pseudobulk, which are used in the tutorials.

Could you please confirm if this module has been renamed, moved, or deprecated in this branch?

Thanks!

---------------------------------------------------------------------------
ModuleNotFoundError                       Traceback (most recent call last)
Cell In[15], line 1
----> 1 from pycisTopic.pseudobulk_peak_calling import export_pseudobulk
      2 os.makedirs(os.path.join(out_dir, "consensus_peak_calling"), exist_ok = True)
      3 os.makedirs(os.path.join(out_dir, "consensus_peak_calling[/pseudobulk_bed_files](http://localhost:8888/pseudobulk_bed_files)"), exist_ok = True)

ModuleNotFoundError: No module named 'pycisTopic.pseudobulk_peak_calling'