Target preference. The model is sequence, ligand, sequence, and the steps below follow that order: the first target, then the compound, then the second target. Add more targets to rank a panel. The two targets can come from different families or from one family, and each answer carries the accuracy measured on its own kind of comparison.
Targets are named by UniProt accession. Gene symbols are many-to-many against accessions and the failure is silent, so the accession is what the model is given and what the results show.
The one compound put to every target.
Pick any second target. If it sits in a different family from the first, the answer carries the across-family figures; if it sits in the same family, it carries that family's own. The page says which it used.
Up to 12 targets per run. Every pair is scored, so n targets is n(n−1)/2 comparisons.
Two targets never needs an email. A longer panel does, and the report is emailed back.
Unlocks the extra target box below ↓
Results go to that address, and we use it to say when the model changes.
Every answer carries one number, its strength, from 0.5 to 1.0, labeled with its band and that band's measured accuracy. The band accuracy comes from the kind of comparison you asked for.
| Strength at or above | Across families | Within one family | ||
|---|---|---|---|---|
| Share kept | Accuracy | Share kept | Accuracy | |
| 0.50, answer everything | 100.0% | 0.75 | 100.0% | 0.78 |
| 0.60 | 70.6% | 0.82 | 76.1% | 0.85 |
| 0.70 | 49.5% | 0.89 | 56.0% | 0.91 |
| 0.80 | 34.3% | 0.94 | 38.3% | 0.95 |
| 0.90 | 20.5% | 0.97 | 20.8% | 0.98 |
A near-tie is a near-tie: where the two measured values sit within half a log the model is right 0.59 of the time across families and 0.64 within one.
The model answers which of two targets a compound prefers. It is not a safety or toxicity screen: breadth across families in this data partly records how many assay panels a compound went through, not how promiscuous it is.