Top candidates discovered from pairwise peak differences. All candidates are pre-selected; uncheck any you want to exclude. At most Max unknown blocks will be used.
Full pairwise biosynthetic network analysis. Lower score = simpler, more plausible network. Click a row to expand. Each node shows its direct-link neighbors (single-step matches) and extra direct links beyond the backbone tree.
| Rank | Model | BIC | Score ↑ | N Good | Direct links | Avg direct links | Main-group coverage | Link groups | Backbone | D/S/M/X | Backbone pen. |
|---|
Inferred biosynthetic dependency tree between blocks based on exhaustive model comparison.
Mass error (observed − reconstructed) vs m/z for each model. Green = GOOD, Red = BAD.
Residual analysis and block usage statistics for each model.