Setting colors for different conditions of cells

[loversaber]

I'm trying to figure out how to display cells with two conditions in two different colors. I don't think there are specific cell labels for each cell input into SPRING because only the X matrix and the gene list are used in the function make_spring_subplot. However, I suppose the indexes from 0 to the total number of cells can be used automatically as cell labels,

I guessed that the build_categ_colors function would be used to make the input for parameter cell_groupings. Then, cell_groupings was set to d_condition_labels as shown in the code below in make_spring_subplot. But the graph generated in SPRING is not what I want. So I don't think I set it up right. Can you help me to solve this problem? Thank you very much!

# test #0-5 indicate the cell label 
# Dict d_condition_labels contains cell labels belonging to the corrsponding condition
d_condition_labels = {"Infected":['3','4','5'],"Control":['0','1','2']}

# Then I tested the function to check the structure of cell_groupings
build_categ_colors({},d_condition_labels)
{'Control': {'label_colors': {'0': '#00d4ff', '1': '#00007f', '2': '#ffe500'},
  'label_list': ['0', '1', '2']},
 'Infected': {'label_colors': {'3': '#00007f', '4': '#ffe500', '5': '#00d4ff'},
  'label_list': ['3', '4', '5']}}

[calebweinreb]

Hello,

Sorry for the lack of documentation. You would indeed using the cell_groupings input of make_spring_plot to color cells by condition. The way to format it is that there should be one key/value pair for partition that you want to visualize (e.g. timepoint, genotype, batch, etc.). In your case, there would just be a single key/value pair because there is only one partition you want to visualize. The key should be the same of the partition (in your case "condition") and the value should be a list of labels, one for each cell in the expression matrix. For example, if cells 0,1,2 are "Control" and cells 3,4,5 are "Infected", the full dictionary would be

cell_groupings = {"Condition": ["Control", "Control", "Control", "Infected", "Infected", "Infected"]}

[loversaber]

Hello,

Sorry for the lack of documentation. You would indeed using the cell_groupings input of make_spring_plot to color cells by condition. The way to format it is that there should be one key/value pair for partition that you want to visualize (e.g. timepoint, genotype, batch, etc.). In your case, there would just be a single key/value pair because there is only one partition you want to visualize. The key should be the same of the partition (in your case "condition") and the value should be a list of labels, one for each cell in the expression matrix. For example, if cells 0,1,2 are "Control" and cells 3,4,5 are "Infected", the full dictionary would be

cell_groupings = {"Condition": ["Control", "Control", "Control", "Infected", "Infected", "Infected"]}

Thank you very much! It really helped me out!