Anti-tumor search
For each drug, the "ideal patient" is the tumor whose pathway state equals −Π_d (cosine similarity = 1). Then we search 9,200 real tumors for the closest match. If even the best one is far, the drug's ideal patient doesn't naturally occur.
Orphan-ideal drugs (0)
No real tumor closer than cosine 0.30 to the drug's ideal patient. Either the disease they were built for is extinct, or we haven't ingested the cohort that contains it yet.
| Drug | cos(closest real, ideal) | Closest tumor |
|---|
Best-matched drugs (top 15)
Drugs whose ideal patient IS well-represented in real tumors. Highest-confidence candidates for the cohort they match.
| Drug | cos(closest real, ideal) | Best match |
|---|---|---|
| Capmatinib | 0.883 | EPT0291 |
| Temsirolimus | 0.881 | EPT0291 |
| Sirolimus | 0.880 | EPT0291 |
| Everolimus | 0.880 | EPT0291 |
| Defactinib | 0.879 | EPT0291 |
| Abemaciclib | 0.873 | EPT0291 |
| Cobimetinib | 0.873 | EPT0291 |
| Asciminib | 0.873 | EPT0291 |
| Palbociclib | 0.872 | EPT0291 |
| Selumetinib | 0.872 | EPT0291 |
| Gedatolisib | 0.869 | EPT0291 |
| Lapatinib | 0.869 | EPT0291 |
| Binimetinib | 0.869 | EPT0291 |
| Leniolisib | 0.868 | EPT0291 |
| Duvelisib | 0.867 | EPT0291 |