Visualization¶

Overview¶

The MS-MINT visualization tools provide powerful and flexible ways to explore mass spectrometry data, offering both static (Matplotlib) and interactive (Plotly) visualization options.

Visualization Types¶

1. Heatmaps¶

Static Heatmaps¶

# Hierarchical clustering heatmap
plotter.hierarchical_clustering(
    data=your_dataframe,      # Data to visualize
    vmin=-3,                  # Minimum value for color scaling
    vmax=3,                   # Maximum value for color scaling
    metric='cosine',          # Distance metric for clustering
    figsize=(8, 8),           # Figure size
    xmaxticks=10,             # Maximum x-axis ticks
    ymaxticks=10              # Maximum y-axis ticks
)

Interactive Heatmaps¶

# Create an interactive heatmap
plotter.heatmap(
    col_name='peak_max',      # Column to visualize
    normed_by_cols=True,      # Normalize by column
    transposed=False,         # Transpose matrix
    clustered=True,           # Apply hierarchical clustering
    correlation=False         # Convert to correlation matrix
)

Heatmap Options¶

Normalization: Scale data by column maximum
Clustering: Hierarchical clustering of rows/columns
Correlation: Convert to correlation matrix
Color Scaling: Customize color range and palette

2. Peak Shapes¶

Static Peak Shape Plots¶

# Plot peak shapes from results
plotter.peak_shapes(
    fns=['file1.mzXML', 'file2.mzXML'],  # Specific files
    peak_labels=['Caffeine', 'Glucose'],  # Specific peaks
    height=3,                 # Facet height
    aspect=1.5,               # Facet aspect ratio
    col_wrap=4,               # Maximum columns
    legend=True               # Show legend
)

Interactive Peak Shape Plots¶

# Interactive peak shape visualization
plotter.peak_shapes(
    fns=['file1.mzXML', 'file2.mzXML'],  # Specific files
    peak_labels=['Caffeine', 'Glucose'],  # Specific peaks
    interactive=True,         # Use Plotly interactive mode
    color='ms_file_label'     # Color by file label
)

3. Chromatograms¶

# Plot chromatograms
plotter.chromatogram(
    fns=['file1.mzXML', 'file2.mzXML'],  # Specific files
    peak_labels=['Caffeine', 'Glucose'],  # Specific peaks
    interactive=False,        # Static plot
    filters=None              # Optional data filters
)

4. 2D Histograms¶

# Create 2D histogram of MS file
plotter.histogram_2d(
    fn='your_ms_file.mzXML',  # MS file to visualize
    peak_label='Caffeine',    # Optional peak label to highlight
    rt_margin=0,              # Retention time margin
    mz_margin=0               # M/Z margin
)

Advanced Visualization Techniques¶

Customization Options¶

Color Palettes: Choose from various color schemes
Interactive vs. Static: Switch between Plotly and Matplotlib
Filtering: Select specific files or peaks
Scaling: Normalize and transform data

Color Customization¶

# Use different color palettes
plotter.peak_shapes(
    palette='Plasma',         # Color palette
    color='ms_file_label'     # Color grouping
)

Best Practices¶

Data Preprocessing: Ensure data is clean and standardized
Choose Appropriate Visualization: Match plot type to your analysis goals
Interactive vs. Static:
Use interactive for detailed exploration
Use static for publications/reports
Color Choices: Select palettes that are color-blind friendly

Troubleshooting¶

No Data Displayed:
Check data filtering
Verify peak labels and file names
Performance Issues:
Reduce data size
Use interactive mode for large datasets
Color Mapping: Ensure unique color mapping for different groups

Performance Tips¶

For large datasets, use interactive Plotly visualizations
Limit the number of peaks and files in a single plot
Use normalization and scaling to improve visualization clarity

Example Workflow¶

from ms_mint import Mint

# Create Mint instance
mint = Mint()

# Load data
mint.load_targets('targets.csv')
mint.load_ms_files('data_directory/*')

# Run analysis
mint.run()

# Create plotter
plotter = mint.plot

# Visualize results
plotter.heatmap(col_name='peak_max')
plotter.peak_shapes(interactive=True)
plotter.chromatogram()

Notes¶

Visualization tools are designed to be flexible and intuitive
Always verify data and visualization parameters
Experiment with different visualization techniques