v0.6.8

Made computation of neighborhoods less memory-intensive.

v0.6.7

Fixed bug in tl.convert_to_anndata().

v0.6.6

Made IPython dependency optional, since the forced install did not work well in environments like Colab.

v0.6.5

What's Changed

• Fix/add labels from dataframe by @MeyerBender in #107
• Quality of life improvements by @MeyerBender in #109

Full Changelog: v0.6.4...v0.6.5

v0.6.4

Feature: pp.add_quantification() now includes an argument called layer_key, with which users can specify which layer to use for quantification. The use case for this is if you have both raw and processed intensity layers in your object.

v0.6.3

This releases includes:

• bugfixes
• updates to the documentation
• features
  • customizable legend order in plots
  • computation of global network measures
  • neighborhood composition visualization
  • possibility to add 3D layers to the object (such as raw data after the '_image' layer has already been processed)

v0.6.2

This release adds network-based neighborhood analysis methods, such as computing node degree, homophily, diversity, etc. Refer to the documentation section on Neighborhood Analysis for more details.

v0.6.1

Refined cell subtype prediction.

1. Double positivity: You can encode a cell type with double positivity. In the case below, we would only call a cell as naive cytotoxic, if both CD8 and CD45RA are positive.

B:
  subtypes:
    - name: B_prol
      markers: ["Ki-67+"]
T:
  subtypes:
    - name: T_h
      markers: ["CD4+"]
    - name: T_tox
      markers: ["CD8+", "CD45RA+"]

2. Alternative markers: Alternatively, you could also define a cell type if either one of a set of markers is positive. Here, cells would be called T_tox if either CD45RO or CD45RA are positive.

B:
  subtypes:
    - name: B_prol
      markers: ["Ki-67+"]
T:
  subtypes:
    - name: T_tox
      markers: ["CD45RA+"]
    - name: T_tox
      markers: ["CD45RO+"]

3. Marker negativity: By using a minus sign after a marker, we call a cell if it is negative for that marker. For example, we could define T helper cells as being CD8 negative.

B:
  subtypes:
    - name: B_prol
      markers: ["Ki-67+"]
T:
  subtypes:
    - name: T_h
      markers: ["CD8-"]
    - name: T_tox
      markers: ["CD8+"]

4. Combined positivity and negativity: We can also use combinations of positive and negative markers to define certain cell subtypes.

B:
  subtypes:
    - name: B_prol
      markers: ["Ki-67+"]
T:
  subtypes:
    - name: T_h
      markers: ["CD4+", "CD8-"]
    - name: T_tox
      markers: ["CD8+"]

v0.6.0

Simplified cell subtype prediction:

# calling of cell subtypes with a provided hierarchy
subtype_dict = {
    'B': {
        'subtypes': [
            {
                'name': 'B_prol',
                'markers': ['Ki67']
            }
        ]
    },
    'T': {
        'subtypes': [
            {
                'name': 'T_h',
                'markers': ['CD4']
            },
            {
                'name': 'T_tox',
                'markers': ['CD8'],
                'subtypes': [
                    {
                        'name': 'T_tox_naive',
                        'markers': ['CD45RA']
                    },
                    {
                        'name': 'T_tox_mem',
                        'markers': ['CD45RO']
                    }
                ]
            }
        ]
    }
}

ds = ds.la.predict_cell_subtypes(subtype_dict)

More details can be found in the documentation.

v0.5.7

Bugfix: merged segmentation masks can now be stored as zarr files.
Deprecated: removed cellpose_denoise() from the tl module because of inconsistencies with the syntax of tl.cellpose().