lasso tool

[adilysyn]

I would like to use the lasso tool on StereoMap to pick up cells that belongs only to a specific cluster.
However, since the lasso pointer to define the region is quite large, I also pick adjacent cells which are from different clusters.
If I choose a cluster choose cells by the lasso, the result will contain also cells that not belong to the cluster?

Thank you in advance,
Adily

[Jinghao-Chen]

Hi, currently, StereoMap does not support selecting cells of a specific cluster via lasso selection. Following your operation method, the result will still include cells that do not belong to the target cluster. We will add support for this feature in the next version. For the time being, we recommend referring to the method provided in this issue to obtain the corresponding cellbin.gef. Once you have obtained the CGEF file for your target cluster, you may then proceed with the Lasso operation on it.
#262

[adilysyn]

Thank you.
Can you please refer to this issue 264 where it was recommended (in the last response) to perform a comparison between two ROI of two experiments, firstly by using the lasso tool. But if the lasso does not support selecting cells of a specific cluster, how can I obtain the comparison between my two ROIs?

[Jinghao-Chen]

1. It is recommended to first export the CGEF files for the specific cluster from both experiments. Refer to the method provided in GEF by cluster #262
2. You may then use StereoMap to perform ROI selection on these two CGEF files output from step 1.
3. Finally, you can run saw reanalyze lasso to generate the corresponding GEF files for comparison.

[adilysyn]

I would like to compare ROIs that composed of several clusters. Should I concatenate CGEFs of different clusters and than upload to StereoMap?

[Jinghao-Chen]

Yes, you may consider using this modified script, which supports inputting multiple clusters separated by commas.

#Please set these parameters
sn=C04042E2 #SN number
cellbin_h5ad=/path/to/outs/analysis/${sn}.cellbin_1.0.adjusted.h5ad
cellbin_gef=/path/to/outs/feature_expression/${sn}.adjusted.cellbin.gef
target_cluster=15,16,17 #the clusters you want to extract
cell_mask=/path/to/outs/image/${sn}_ssDNA_mask_edm_dis_10.tif #If you intend to process the cellbin.gef rather than adjusted.cellbin.gef, please set it to /path/to/outs/image/${sn}_ssDNA_mask.tif
saw_path=/path/to/saw_path #e.g. /path/to/saw-8.2.1
out_path=/path/to/output_directory
raw_gef=/path/to/outs/feature_expression/${sn}.raw.gef

#Then run the attached script get_specific_cluster_cgef_v2.py
${saw_path}/anaconda/bin/python get_specific_cluster_cgef.py ${cellbin_h5ad} ${cellbin_gef} ${target_cluster} ${cell_mask} ${saw_path} ${out_path} ${raw_gef}

The output cellbin.gef file will be saved to ${out_path}.

[adilysyn]

Thank you very much!

[Jinghao-Chen]

@adilysyn My apologies, but a bug was found in the script I shared earlier. Should you need to use the script, please kindly replace it with this updated one:

get_specific_cluster_cgef_v2.py

Best wishes!
Jinghao

[adilysyn]

Thank you, I'll use the updated one.

[adilysyn]

[41][INFO]: start to run filter_cells...
[2026-02-15 06:42:47][Stereo][2748039][MainThread][126494132344640][st_pipeline][44][INFO]: filter_cells end, consume time 0.3392s.
Processing labels:  10%|█▍             | 35636/364094 [57:39<9:28:59,  9.62it/s]

There is a way to make it run faster? only 10% after an hour. I have 64 cpu cores.

[adilysyn]

Another, more critical question: I used the get_specific_cluster_cgef_v2.py script to generate CGEF files for several clusters and uploaded them to StereoMap. However, when selecting the “Cluster” layer, no clusters are visible. When selecting the “Heatmap” layer, only some unrelated clusters appear.

The purpose is to perform ROI selection on the CGEF file, but this is currently not possible.

Could you please help determine why the script output does not match the original clustering (as seen in the source data)?

[Jinghao-Chen]

The clustering information was not provided in the previous script. You could use the following script to add the corresponding clustering information, and input the cellbin.gef file you obtained by running get_specific_cluster_cgef_v2.py.

add_cluster_info.py

#Please set these parameters
sn=C04042E2 #SN number
cellbin_h5ad=/path/to/outs/analysis/${sn}.cellbin_1.0.adjusted.h5ad
cellbin_gef={cellbin.gef obtained by running get_specific_cluster_cgef_v2.py}

saw_path=/path/to/saw_path #e.g. /path/to/saw-8.2.1


#Then run the attached script add_cluster_info.py
${saw_path}/anaconda/bin/python add_cluster_info.py ${cellbin_h5ad} ${cellbin_gef} ${saw_path}

The cellbin.gef file will be updated.

If you find that the clustering results are still inconsistent, you could send these files to me for troubleshooting.

/path/to/outs/analysis/${sn}.cellbin_1.0.adjusted.h5ad
/path/to/outs/feature_expression/${sn}.adjusted.cellbin.gef
target_cluster #the clusters you want to extract
/path/to/outs/image/${sn}_ssDNA_mask_edm_dis_10.tif 
/path/to/outs/feature_expression/${sn}.raw.gef

[adilysyn]

I run the script add_cluster_info.py with the appropriate arguments, and got this error:

Traceback (most recent call last):
  File "/home/adily/STOmics/saw/package/saw-8.2.2/anaconda/lib/python3.8/site-packages/pandas/core/indexes/base.py", line 3802, in get_loc
    return self._engine.get_loc(casted_key)
  File "pandas/_libs/index.pyx", line 138, in pandas._libs.index.IndexEngine.get_loc
  File "pandas/_libs/index.pyx", line 165, in pandas._libs.index.IndexEngine.get_loc
  File "pandas/_libs/hashtable_class_helper.pxi", line 5745, in pandas._libs.hashtable.PyObjectHashTable.get_item
  File "pandas/_libs/hashtable_class_helper.pxi", line 5753, in pandas._libs.hashtable.PyObjectHashTable.get_item
KeyError: '40604620819698'

The above exception was the direct cause of the following exception:

Traceback (most recent call last):
  File "add_cluster_info.py", line 20, in <module>
    new_cluster_ids.append(int(adata.obs.loc[cell_id, 'leiden']))
  File "/home/adily/STOmics/saw/package/saw-8.2.2/anaconda/lib/python3.8/site-packages/pandas/core/indexing.py", line 1066, in __getitem__
    return self.obj._get_value(*key, takeable=self._takeable)
  File "/home/adily/STOmics/saw/package/saw-8.2.2/anaconda/lib/python3.8/site-packages/pandas/core/frame.py", line 3924, in _get_value
    row = self.index.get_loc(index)
  File "/home/adily/STOmics/saw/package/saw-8.2.2/anaconda/lib/python3.8/site-packages/pandas/core/indexes/base.py", line 3804, in get_loc
    raise KeyError(key) from err
KeyError: '40604620819698'

If the previous script - get_specific_cluster_cgef_v2.py - doesn't provide the clustering information, what it does?

[Jinghao-Chen]

The script provided earlier only extracts the expression matrix of cells in the corresponding clusters, but does not record the cluster information. Since the current version of StereoMap does not support this function yet, you may provide these files and the clusters you are interested in if convenient., and I will help you with the extraction.

[adilysyn]

Thank you for the clarification.

By “help with the extraction,” do you mean selecting the ROI for the clusters of interest? We are working with several experiments, each containing multiple clusters, and we need to define ROIs for each cluster.

Is there currently any way for us to perform this extraction ourselves? If not, how should we provide you with the relevant files and cluster information?

Thank you for your help,
Adily