Available plots in Owkin K

Owkin K will automatically select the most appropriate visualization type based on your question if you don't specify one, but being explicit about the plot type you want can help ensure you get exactly the visualization you need.

This page provides comprehensive documentation for all plot tools available in the Owkin K. The plots are organized by data modality. Each plot is designed for specific analysis scenarios, from differential expression analysis to spatial colocalization, survival analysis, and treatment timeline visualization.

Plots support various parameters for customization including patient grouping or filtering, and visualization options.

How to Request Visualizations

When requesting visualizations in Owkin K, you can:

1. Ask for a specific plot type by name (e.g., "Generate a Kaplan-Meier plot...")

2. Describe what you want to visualize without specifying the plot type (e.g., "Show me the relationship between...")

3. Request multiple visualizations in a single query (e.g., "Compare survival and gene expression...")

Summary

Plots overview

This documentation covers 28 different plot types across 6 data modalities:

• Clinical: 3 plots (1 available in K Pro Free)

• Bulk RNA-Seq: 6 plots (5 available in K Pro Free, 1 requires K Pro)

• Single-Cell RNA-Seq: 7 plots (6 available in K Pro Free, 1 requires K Pro)

• Spatial Transcriptomics: 7 plots (5 available in K Pro Free, 2 require K Pro)

• Histomics: 1 plot (1 available in K Pro Free)

• Proteomics: 4 plots (0 available in K Pro Free, 4 require K Pro)

K Pro Free Availability

Each plot includes a K Pro indicator showing its availability:

• ✅ K Pro Free - Available in K Pro Free

• ⭐ K Pro only - Unlock with K Pro

K Pro Free Availability Summary:

• ✅ 18 plots are available in K Pro Free

• ⭐ 10 plots require a K Pro subscription

Plot Types

Owkin K offers a variety of specialized visualization types designed for biomedical research. Below is a comprehensive list of available plots that you can request in your queries:

Clinical Outcome Plots

Kaplan-Meier Plot (clinical_endpoint_plot)

• Visualizes survival analysis with time-to-event data

• Includes log-rank test p-values and hazard ratios

• Can be stratified by various factors (mutation status, expression levels, etc.)

• Example query: "Generate a Kaplan-Meier plot comparing overall survival between TP53 mutated and wildtype lung cancer patients"

Molecular Data Plots

1. Heatmap (expression_heatmap)

   • Displays gene expression patterns across multiple samples

   • Uses color intensity to represent expression levels

   • Can include hierarchical clustering of genes and/or samples

   • Example query: "Create a heatmap of the top 50 most variable genes in breast cancer samples"

2. Volcano Plot (differential_expression_plot)

   • Shows statistical significance versus magnitude of change

   • Highlights differentially expressed genes between conditions

   • Example query: "Generate a volcano plot comparing gene expression between responders and non-responders"

3. Box Plot / Violin Plot (expression_plot)

   • Compares distribution of expression values across groups

   • Shows median, quartiles, and outliers

   • Example query: "Create box plots showing EGFR expression across different lung cancer subtypes"

1. Mutation Oncoprint

   • Visualizes genetic alterations across patient cohorts

   • Shows mutation types, frequencies, and co-occurrence patterns

   • Example query: "Generate an oncoprint showing mutation patterns in key driver genes across colorectal cancer patients"

Relationship Plots

1. Correlation Plot (correlation_plot)

   • Displays correlation coefficients between multiple variables

   • Uses color intensity to represent correlation strength

   • Example query: "Create a correlation plot for expression of immune checkpoint genes in melanoma"

2. Scatter Plot (scatter_plot)

   • Shows relationship between two continuous variables

   • Can include regression lines and confidence intervals

   • Example query: "Generate a scatter plot of tumor mutational burden versus immune infiltration score"

3. Network Diagram

   • Visualizes interactions between genes, proteins, or other entities

   • Shows connection patterns and node importance

   • Example query: "Create a network diagram of protein-protein interactions for the p53 pathway"

4. Dimensionality Reduction Plot (pca_plot, tsne_plot, umap_plot)

   • Reduces high-dimensional data to 2D or 3D for visualization

   • Options include PCA, t-SNE, and UMAP

   • Example query: "Generate a PCA plot of gene expression data colored by cancer subtype"

Patient Cohort Plots

1. Clinical Summary Table

   • Presents demographic and clinical characteristics of patient cohorts

   • Shows counts and percentages for categorical variables

   • Displays means, medians, and ranges for continuous variables

   • Example query: "Create a clinical summary table for breast cancer patients in the TCGA dataset"

2. Bar Chart (bar_plot)

   • Compares frequencies or values across categories

   • Can be grouped or stacked for multi-level comparisons

   • Example query: "Generate a bar chart showing mutation frequencies in key driver genes across cancer types"

3. Pie Chart (pie_chart)

   • Shows proportion of different categories within a whole

   • Example query: "Create a pie chart of cancer stage distribution in lung adenocarcinoma patients"

4. Sankey Diagram

   • Visualizes flows between categories

   • Useful for showing patient treatment pathways or transitions between states

   • Example query: "Generate a Sankey diagram showing treatment sequences in metastatic breast cancer patients"