Skip to content

Sm.pl.foldchange

Short Description

sm.pl.foldchange: The Function allows users to visualize foldchange in abundance of celltypes between samples/ROI's. Run sm.tl.foldchange first to compute the foldchange.

Function

foldchange(adata, label='foldchange', p_val=0.05, nonsig_color='grey', subset_xaxis=None, subset_yaxis=None, cmap='vlag', log=True, center=0, method='heatmap', invert_axis=None, parallel_coordinates_color=None, matplotlib_bbox_to_anchor=(1.04, 1), matplotlib_legend_loc='upper left', xticks_rotation=90, return_data=False, **kwargs)

Parameters:

Name Type Description Default
adata

Anndata object

required
label

strong, optional
label used when running sm.tl.foldchange.

'foldchange'
p_val

float, optional
p_val cut-off above which is considered not-significant. The cells containing non-significant changes will be highlighted in the heatmap.

0.05
nonsig_color

string, optional
Color used to highlight non-significant fold changes in the heatmap.

'grey'
subset_xaxis

list, optional
Subset x-axis before plotting. Pass in a list of categories. eg- subset_xaxis = ['CelltypeA', 'CellTypeB'].

None
subset_yaxis

list, optional
Subset y-axis before plotting. Pass in a list of categories. eg- subset_yaxis = ['ROI_1', 'ROI_5'].

None
cmap

string, optional
Color map. Can be a name or a Colormap instance (e.g. 'magma', 'viridis').

'vlag'
log

bool, optional
Convert foldchange to log2 scale.

True
center

float, optional
The center value to be used in heatmap.

0
method

string, optional
Two methods are available for plotting the foldchanges
a) Heatmap: Use heatmap
b) parallel coordinates plot : Use parallel_coordinates

'heatmap'
invert_axis

bool, optional
Flip the axis of the plot.

None
parallel_coordinates_color

list, optional
Custom colors for each category.

None
matplotlib_bbox_to_anchor

tuple, optional
Bounding box argument used along with matplotlib_legend_loc to control the legend location when using the matplotlib method.

(1.04, 1)
matplotlib_legend_loc

TYPE, optional
Location of legend used along with matplotlib_bbox_to_anchor to control the legend location when using the matplotlib method.

'upper left'
xticks_rotation

int, optional
Angle the x-axis ticks.

90
return_data

bool, optional
Return the final data used for plotting.

False
**kwargs

Additional keyword arguments passed to:
a) sns.clustermap
b) pandas.parallel_coordinates

{}

Returns:

Name Type Description
Plot

Data used for the plot if return_data = True

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
    # Heatmap of foldchnage  
    sm.pl.foldchange (adata, label='foldchange', method='heatmap',
                     p_val=0.05, nonsig_color='grey',
                     cmap = 'vlag', log=True, center=0, linecolor='black',linewidths=0.7,
                     vmin=-5, vmax=5, row_cluster=False)

    # Parallel_coordinates plot of the foldchanges
    foldchange (adata, label='foldchange', 
                log=True, method='parallel_coordinates', invert_axis=True,
                parallel_coordinates_color=['black','blue','green','red','#000000'],
                matplotlib_bbox_to_anchor=(1.04,1),
                matplotlib_legend_loc='upper left',
                xticks_rotation=90,
                return_data = False
Source code in scimap/plotting/_foldchange.py
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
def foldchange (adata, label='foldchange', 
                p_val=0.05, nonsig_color='grey',subset_xaxis=None,subset_yaxis=None,
                cmap = 'vlag', log=True,center=0, 
                method='heatmap', invert_axis=None,
                parallel_coordinates_color=None,matplotlib_bbox_to_anchor=(1.04,1),
                matplotlib_legend_loc='upper left',xticks_rotation=90,
                return_data = False,
                **kwargs):
    """
Parameters:
    adata : Anndata object

    label : strong, optional  
        label used when running `sm.tl.foldchange`.

    p_val : float, optional  
        p_val cut-off above which is considered not-significant. The cells containing
        non-significant changes will be highlighted in the heatmap.

    nonsig_color : string, optional  
        Color used to highlight non-significant fold changes in the heatmap.

    subset_xaxis : list, optional  
        Subset x-axis before plotting. Pass in a list of categories. eg- subset_xaxis = ['CelltypeA', 'CellTypeB']. 

    subset_yaxis : list, optional  
        Subset y-axis before plotting. Pass in a list of categories. eg- subset_yaxis = ['ROI_1', 'ROI_5']. 

    cmap : string, optional  
        Color map. Can be a name or a Colormap instance (e.g. 'magma', 'viridis').

    log : bool, optional  
        Convert foldchange to log2 scale.

    center : float, optional  
        The center value to be used in heatmap.

    method : string, optional  
        Two methods are available for plotting the foldchanges  
        a) Heatmap: Use `heatmap`  
        b) parallel coordinates plot : Use `parallel_coordinates`  

    invert_axis : bool, optional  
        Flip the axis of the plot.

    parallel_coordinates_color : list, optional  
        Custom colors for each category.

    matplotlib_bbox_to_anchor : tuple, optional  
        Bounding box argument used along with matplotlib_legend_loc to control
        the legend location when using the matplotlib method.

    matplotlib_legend_loc : TYPE, optional  
        Location of legend used along with matplotlib_bbox_to_anchor to control
        the legend location when using the matplotlib method.

    xticks_rotation : int, optional  
        Angle the x-axis ticks.

    return_data: bool, optional  
        Return the final data used for plotting.

    **kwargs : Additional keyword arguments passed to:  
        a) sns.clustermap  
        b) pandas.parallel_coordinates  

Returns:
    Plot:  
        Data used for the plot if `return_data = True`

Example:
```python
    # Heatmap of foldchnage  
    sm.pl.foldchange (adata, label='foldchange', method='heatmap',
                     p_val=0.05, nonsig_color='grey',
                     cmap = 'vlag', log=True, center=0, linecolor='black',linewidths=0.7,
                     vmin=-5, vmax=5, row_cluster=False)

    # Parallel_coordinates plot of the foldchanges
    foldchange (adata, label='foldchange', 
                log=True, method='parallel_coordinates', invert_axis=True,
                parallel_coordinates_color=['black','blue','green','red','#000000'],
                matplotlib_bbox_to_anchor=(1.04,1),
                matplotlib_legend_loc='upper left',
                xticks_rotation=90,
                return_data = False
```
    """

    # set color for heatmap
    #cmap_updated = copy.copy(matplotlib.cm.get_cmap(cmap))
    cmap_updated = matplotlib.cm.get_cmap(cmap)
    cmap_updated.set_bad(color=nonsig_color)


    # get the data
    fc = adata.uns[str(label)+'_fc']
    p = adata.uns[str(label)+'_pval']

    #fold
    fold = fc.copy()
    p_mask = p.copy()

    # reference image
    ref = fold.index.name

    # log
    if log is True:
        fold = np.log2(fold)

    # create a mask for non-sig values
    p_mask[p_mask > p_val] = np.nan

    # subset x axis data
    if subset_xaxis is not None:
        if isinstance (subset_xaxis, str):
            subset_xaxis = [subset_xaxis]
        fold = fold [subset_xaxis]
        p_mask = p_mask [subset_xaxis]
        #reorder

    # subset y axis data
    if subset_yaxis is not None:
        if isinstance (subset_yaxis, str):
            subset_yaxis = [subset_yaxis]
        fold = fold.loc [subset_yaxis]
        p_mask = p_mask.loc [subset_yaxis]
        #reorder

    # invert axis if user requests
    if invert_axis is True:
        fold = fold.T
        p_mask = p_mask.T

    #mask
    mask = p_mask.isnull() # identify the NAN's for masking 

    if method == 'heatmap':
        # heatmap of the foldchange
        #g= sns.clustermap(fold, cmap=cmap, mask=mask, center=center, col_cluster=False, row_cluster=False)
        g= sns.clustermap(fold, cmap=cmap, mask=mask, center=center, **kwargs)
        plt.suptitle('reference: '+ str(ref))
        plt.setp(g.ax_heatmap.get_xticklabels(), rotation=xticks_rotation)
        plt.tight_layout()


    if method == 'parallel_coordinates':
        fold['sample'] = fold.index
        # plotting
        fig, axes = plt.subplots()
        if parallel_coordinates_color is not None:
            parallel_coordinates(fold, 'sample', color=parallel_coordinates_color, **kwargs)
        else:
            #parallel_coordinates(fold, 'sample', colormap=cmap_updated)
            parallel_coordinates(fold, 'sample', colormap=cmap_updated, **kwargs)
        axes.grid(False)
        plt.legend(bbox_to_anchor=matplotlib_bbox_to_anchor, loc=matplotlib_legend_loc)
        plt.axhline(y=0, color='black', linestyle='-')
        plt.xticks(rotation = xticks_rotation)
        plt.suptitle('reference: '+ str(ref))
        fig.tight_layout()

    # return data
    if return_data is True:
        return fold