Skip to main content

Visualisation Functions (Python)

This page lists the Statescope visualisation helpers and when to use them. Most plots expect that you have run the relevant module (Deconvolution, Refinement, or StateDiscovery) first.

Quick start​

from Statescope import Initialize_Statescope
from Statescope import (
    Heatmap_Fractions,
    Boxplot_CelltypeAbundance,
    Heatmap_GEX,
    Heatmap_StateScores,
    Heatmap_StateLoadings,
    Plot_CopheneticCoefficients,
    BarPlot_StateLoadings,
    TSNE_AllStates,
    TSNE_CellTypes,
)

Statescope_model = Initialize_Statescope(Bulk, TumorType="PBMC", Ncelltypes=7, Ncores=10)
Statescope_model.Deconvolution()
Statescope_model.Refinement()
Statescope_model.StateDiscovery()

Deconvolution outputs​

Heatmap_Fractions​

Visualizes cell-type fractions across samples.

Heatmap_Fractions(Statescope_model)

Heatmap Fractions

Boxplot_CelltypeAbundance​

Boxplot of cell-type fractions across samples.

Boxplot_CelltypeAbundance(Statescope_model)

Heatmap Fractions

Refinement outputs​

Heatmap_GEX​

Clustered heatmap of purified gene expression for a cell type.

Heatmap_GEX(Statescope_model, "T_cells_CD4+")

Heatmap Fractions​

State discovery outputs​

Heatmap_StateScores​

Heatmap of per-sample state scores.

Heatmap_StateScores(Statescope_model)

Heatmap Fractions

Heatmap_StateLoadings​

Heatmap of state loadings; use top_genes to reduce clutter.

Heatmap_StateLoadings(Statescope_model, top_genes=50)

Heatmap Fractions

Plot_CopheneticCoefficients​

Plots cophenetic coefficients across k values (used for model selection).

Plot_CopheneticCoefficients(Statescope_model)

Heatmap Fractions

BarPlot_StateLoadings​

Bar plot of state loadings with top genes labeled.

BarPlot_StateLoadings(Statescope_model, top_genes=3)

Heatmap Fractions

TSNE_AllStates​

t-SNE for all cell types and states (optionally show sample names).

TSNE_AllStates(Statescope_model, show_samples=True)

TSNE_CellTypes​

t-SNE for one or more cell types.

TSNE_CellTypes(Statescope_model, celltype="Epithelial", show_samples=True)

Usage notes​

  • Run Deconvolution() before fraction plots.
  • Run Refinement() before GEX heatmaps.
  • Run StateDiscovery() before state score/loading plots and t-SNE views.
  • It is recommended to save your model after each module so you can reload and continue analysis later.