Plot

deside.plot.plot_corr_two_columns(df: DataFrame, output_dir: str, col_name1: str = 'CPE', col_name2: str = 'cancer_cell', cancer_type: str = '', diagonal: bool = True, predicted_by: str | None = None, font_scale: float = 1.5, scale_exp=False, update_figures=False, cell_type2subtypes: dict | None = None)[source]

Plot the relation between two columns in DataFrame df

Parameters:

  • df – a dataFrame which contains CPE (cancer purity) and cancer_fraction

  • output_dir – result folder

  • col_name1 – column name, such as CPE (cancer purity), x axis

  • col_name2 – column name, such as cancer cell fraction (predicted cancer purity), y axis

  • cancer_type – mark x axis / y axis label

  • diagonal – if plot diagonal

  • predicted_by – model name

  • font_scale – scale font size

  • scale_exp – if scale all expression values to range [0, 10] by x_i/max(x) * 10

  • update_figures – if update figures in output_dir

  • cell_type2subtypes – dict, cell type to subtypes, such as {‘B cells’: [‘B cells naive’, ‘B cells memory’]}

Returns:

None

deside.plot.plot_predicted_result(cell_frac_result_fp, bulk_exp_fp, cancer_type, model_name, result_dir, cancer_purity_fp: str | None = None, font_scale=2.0, update_figures=False, cell_type2subtypes=None)[source]

Plot and evaluate predicted results of DeSide or Scaden model for TCGA data

Parameters:

  • cell_frac_result_fp – the file path of predicted cell fraction

  • bulk_exp_fp – the file path of bulk cell expression profile or pd.Dataframe, TPM, gene by sample

  • cancer_type – only for naming or mark x / y label when plotting

  • model_name – model name, DeSide or Scaden

  • result_dir – where to save result

  • cancer_purity_fp – estimated tumor purity for TCGA, download from Aran, D. et al., Nat Commun 6, 8971 (2015), Supplementary Data 1

  • font_scale – scale font size

  • update_figures – whether to update figures

  • cell_type2subtypes – dict, cell type to subtypes, e.g. {‘CD8 T’: [’…’, ‘…’], }

Returns:

None