Visualize

The Visualize dropdown provides statistical charts derived from the processed targeted dataset. Each chart is rendered with Plotly and supports interactive zoom, hover, and PNG export.

[Screenshot: PCA plot colored by sample groups]

Open a visualization

Click the Visualize tab dropdown in the top action bar. Pick a chart type from the menu.

Available charts

Chart	What it shows	Best for
PCA	Principal component analysis in 2D or 3D	Sample clustering, outlier detection
Correlation	Sample-to-sample correlation matrix heatmap	QC — replicates should correlate
Heatmap	Metabolite intensity matrix with hierarchical clustering	Pattern discovery across many compounds
Box Plot	Distribution of intensities by sample or metabolite	Spotting outliers per compound
Dendrogram	Hierarchical clustering tree	Sample or metabolite relationships
Volcano	Log2 fold-change vs −log10 p-value	Differential analysis between two groups
QC Dashboard	Detection rates + verdict distribution + scoring summary	Reviewing run quality before interpretation
Metabolite Network	Correlation-based network of co-regulated metabolites	Module detection
Sample Network	Similarity network of samples	Checking whether replicates cluster together
Bipartite Network	Sample-to-metabolite connection graph	Top contributors per sample
Pathway Network	KEGG pathway mapping	Biological context

Settings sidebar

Each chart has a collapsible Settings sidebar on the right with chart-specific options. Common ones:

Setting	Where it appears	What it does
Normalization	Heatmap, PCA, networks	Z-score, log2, raw, or median centering
Clustering	Heatmap, dendrogram	Linkage method (ward, complete, average) and metric (euclidean, correlation)
Group coloring	Most charts	Color points by the sample groups defined in the sidebar
Threshold sliders	Volcano, networks	Fold-change cutoff, p-value cutoff, edge threshold
2D / 3D	PCA	Toggle dimensionality

Volcano plot — special requirements

Volcano needs exactly two sample groups. Set them up first:

1. Open the sample selector sidebar from the Peak Picking view
2. Use the lightning-bolt button to auto-group, or manually assign samples to two groups (e.g., WT and KO)
3. Switch to Visualize → Volcano
4. Configure the fold-change and p-value cutoffs in the sidebar

Compounds passing both thresholds appear as colored points. Hover any point for the compound name and stats.

[Screenshot: volcano plot with significant compounds highlighted]

Network charts

The four network types use different inputs:

• Metabolite network — edges are pairwise correlations between metabolites
• Sample network — edges are pairwise correlations between samples
• Bipartite network — edges connect samples to their top contributing metabolites
• Pathway network — edges come from KEGG pathway membership

Adjust the edge threshold slider to control density. Drag nodes to rearrange.

Exporting charts

Plotly's built-in toolbar (top-right of each chart) has:

• 📷 Camera — download as PNG
• 🔍 Zoom, pan, reset
• 📏 Box select — for downstream filtering

For vector output (SVG / PDF), use the browser's print-to-PDF function. Plotly's built-in image export currently produces raster PNG only.

Also from a script

The visualizations on this page are UI-only — generated by Plotly in the browser from the data in a .msd. To produce the same charts headlessly, load the .msd from Python and use the matching matplotlib / plotly pattern:

from leaf.analyzer import Samples
samples = Samples.load("analysis.msd")
compounds_df = samples.compounds_list   # pandas DataFrame for downstream plotting

→ Python recipes

Next step

→ Export results