Dynamic Web Tool¶
The VirMolAnalyte Dynamic Web Tool provides an online workflow for natural product database screening and fragment-level attribution from 13C/DEPT NMR peak lists.
What It Does¶
- Database Search: enter
ppm, q/t/d/speaks, choose a database and evaluator, then review ranked candidate structures. - Spectrum Comparison: click a candidate to compare the experimental spectrum with database-predicted shifts.
- Fragment Attribution: run masking AAS attribution to identify structural regions that support the candidate assignment.
- Structure Comparison: inspect predicted chemical shifts and matched experimental ppm values on paired structure views.
- Fragment Evidence: review each positive fragment with its structure, matched peaks, and spectrum assignment.
Input Format¶
Paste one peak per line:
131.2, s
116.3, d
36.8, t
27.1, q
The accepted carbon-type labels are q, t, d, and s. Labels such as
CH3, CH2, CH, and C are also accepted.
Attribution Parameters¶
The default settings are suitable for exploratory analysis. Increase iterations when a more stable attribution map is needed.
| Group | Main controls |
|---|---|
| Masking sampling | iterations, random seed, connected subgraph size, mask fraction |
| Matching constraints | DEPT type constraint, greedy unique matching |
| Fragment extraction | score threshold, minimum carbons, bridge carbons, max fragments |
| Rendering | fixed color range, atom index labels, predicted-shift labels, assignment max error |
By default, spectrum-assignment links are drawn only when the absolute
chemical-shift difference is <= 5 ppm.
Result Views¶
The result panel combines several complementary views:
- full candidate structure attribution heatmap
- experimental/database spectrum comparison
- spectrum assignment plot with fragment-to-peak links
- predicted-vs-matched structure comparison
- fragment cards with structure and spectrum-assignment evidence
Use these views together: the ranked hit gives the candidate, the spectrum comparison checks global agreement, and the fragment attribution highlights which substructures explain the observed peaks.