how to generate a tumor model?

[zhangyafeng1]

1. model_column = 'Tumor_model_annot'
2. samples = data_annot.loc[data_annot['Tumor_model_annot'] == 'cancer_cells'].index
3. cancer_expr = data_expr[samples]
4. cancer_annot = data_annot.loc[samples]
5. cancer_annot['Tumor_type'] = cancer_annot['Dataset']
6. cancer_annot = cancer_annot[['Tumor_type', 'Dataset']]
8. samples = data_annot.loc[~data_annot[model_column].isna() & (data_annot['Tumor_model_annot'] != 'cancer_cells')].index
10. cells_expr = data_expr[samples]
11. cells_annot = data_annot.loc[samples]
12. cells_annot = cells_annot[[model_column, 'Dataset']]
13. cells_annot.columns = ['Cell_type', 'Dataset']
14. cells_annot = pd.concat([lab_annot, cells_annot])
15. cells_annot.loc[cells_annot['Dataset'].isna(), 'Dataset'] = cells_annot.loc[cells_annot['Dataset'].isna()].index
16. cells_expr = pd.concat([lab_expr, cells_expr], axis=1)

18. to make sure that there is no repeated samples

19. samples = sorted(list(set(cells_annot.index).intersection(set(cells_expr.columns))))
20. cells_expr = cells_expr[samples]
21. cells_annot = cells_annot.loc[samples]
23. print(cells_expr.shape, cells_annot.shape)
24. print(cancer_expr.shape, cancer_annot.shape)

26. adding missing cell types

27. cell_types = CellTypes.load('configs/cell_types.yaml')
28. missing_cts = [x for x in cell_types.get_all_subtypes('General_cells') if not x in cells_annot['Cell_type'].unique()]
30. for ct in missing_cts:

31. subtypes = cell_types.get_direct_subtypes(ct)
    

32. annot = cells_annot.loc[cells_annot['Cell_type'].isin(subtypes)]
    

33. annot.index
    

34. expr = cells_expr[annot.index]
    

35. annot['Cell_type'] = ct
    

36. annot.index = annot.index + f'_{ct}'
    

37. annot['Dataset'] = annot.index
    

38. expr.columns = expr.columns + f'_{ct}'
    

39. cells_expr = pd.concat([cells_expr, expr], axis=1)
    

40. cells_annot = pd.concat([cells_annot, annot])
    

42. to make sure that there is no repeated samples

43. samples = sorted(list(set(cells_annot.index).intersection(set(cells_expr.columns))))
44. cells_expr = cells_expr[samples]
45. cells_annot = cells_annot.loc[samples]
46. print(cells_expr.shape, cells_annot.shape)

49. Model training

50. mixer = Mixer(cell_types=cell_types,

51.           cells_expr=cells_expr, cells_annot=cells_annot,
    

52.           tumor_expr=cancer_expr, tumor_annot=cancer_annot,
    

53.           num_av=3, num_points=300000)
    

55. model = DeconvolutionModel(cell_types,

56.                        boosting_params_first_step='configs/boosting_params/lgb_parameters_first_step.tsv',
    

57.                        boosting_params_second_step='configs/boosting_params/lgb_parameters_second_step.tsv')
    

59. model.fit(mixer)

Hello, I want to generate a tumor model. In the above code, I changed "Blood_model_annot" in line 1 to "Tumor_model_annot", made modifications in line 8, and changed line 27 to "cell_types.yaml". I didn't make any changes elsewhere. Is this acceptable? After execution, an error occurred at the step of model.fit(mixer), as shown in the figure. Could you please tell me the reason and how I should modify it?