Dashboard Guide
Dashboard Guide
Overview
The Oxonium Browser dashboard is an interactive web application served at http://localhost:8051 after a successful pipeline run. It provides real-time filtering, multiple visualization types, and export functionality for exploring oxonium ion detection results. This page walks through each component from top to bottom.
Filtering Controls
Three threshold inputs at the top of the dashboard control which oxonium ions are displayed across all visualizations:
- Counts — minimum number of MS2 spectra where the ion was detected (default: 20)
- Spectral intensity — minimum average normalized intensity as percentage of total spectrum intensity (default: 0.2%)
- Spectral presence — minimum percentage of all MS2 spectra containing the ion (default: 0.02%)
Adjusting any threshold immediately updates the match table, checklist options, and co-occurrence plot. For guidance on choosing appropriate thresholds, see Detection Metrics.
Database Toggle
When chemspace search is enabled (CHEMSPACE_SEARCH=True), a radio button appears below the thresholds:
- Curated — show only results from the user-provided sugar database with its own test masses
- Chemspace — show only results from the chemical space database with its own test masses
- Both — show all results combined
Switching to chemspace or both automatically adjusts the default thresholds to stricter values (counts ≥ 25, intensity ≥ 1.0%, presence ≥ 0.025%) to manage the larger number of candidates. Switching back to curated restores the original defaults. For more details, see the chemspace section in Sugar Database.
Match Table
All oxonium ions passing the current thresholds are listed in a sortable table with the following columns:
| Column | Description |
|---|---|
| group | Water loss family ID — ions differing by ~18 Da are grouped together |
| Oxonium | Sugar name (curated) or molecular formula (chemspace) |
| mass (Da) | Primary diagnostic mass (ox_mass1) |
| count | Number of spectra with positive detection |
| intensity % | Average normalized intensity across positive detections |
| presence % | Percentage of all spectra containing the ion |
Water Loss Grouping
Oxonium ions whose masses differ by approximately 18.011 Da (one water loss) or 36.021 Da (two water losses) are assigned to the same group. For example, Hex (163.06), Hex−H₂O (145.05), and Hex−2H₂O (127.04) would appear as a block.
Color Coding
Metric columns use color gradients scaled by magnitude: blue for counts, green for intensity, red for presence.
Sorting and Export
Click any column header to sort the table. The export table button below the table exports the current filtered view to Excel.
Clustered Co-occurrence Heatmap
Displays pairwise co-occurrence between the checklist-selected oxonium ions. Select at least 2 ions to activate.
What the cells show (co-occurrence)
Each cell shows the fraction of ion A's scans that also contain ion B. The matrix is asymmetric — cell (A, B) can differ from cell (B, A) if the two ions have different total scan counts. Only scans above the current intensity threshold contribute.
How rows and columns are ordered (Jaccard clustering)
Jaccard similarity (|A∩B| / |A∪B|, symmetric) is computed for all pairs and used for hierarchical clustering (average linkage). The resulting dendrogram is drawn to the left of the y-axis labels, showing which ions have the most similar scan profiles.
Interpreting the plot
- Dark diagonal blocks — groups of ions that frequently appear in the same spectra, likely fragments from the same glycan
- Isolated ions with low co-occurrence — may be independent sugar modifications or potential false positives
- Dendrogram branches — short branches connecting ions indicate high similarity; long branches indicate distinct scan profiles
Retention Time Profiles
An extracted ion chromatogram (XIC) plot showing the retention time distribution of selected oxonium ions. Select ions using the checklist below the plot — each checked ion appears as a separate trace.
Useful for:
- Verifying that an oxonium ion elutes across a chromatographic region consistent with glycopeptides
- Identifying false positives that appear only at the beginning or end of the run
- Checking whether related sugars from the same glycan co-elute
Mass Error Distribution
Histogram plots showing the distribution of mass errors for each amino acid fragment reference peak used during two-pass recalibration. Dashed lines mark the 5th and 95th percentile bounds. The title reports the average mass error for 90% of masses.
Use this plot to:
- Verify that recalibration improved mass accuracy
- Determine an appropriate MASS_ERROR parameter value for your data (see Detection Parameters)
