Mesenchymal Reversal Workbench
Combination designer — minimal set-cover of the mesenchymal program
The greedy weighted set-cover finds the smallest drug combination whose members cover complementary mesenchymal kinases to strong inhibition, prioritising the most mesenchymal-enriched kinases (Table S8 fold-change). No single inhibitor covers the whole program — this is where a combination beats monotherapy. Adjust the threshold below.
| # | Add drug | New kinases | Newly covered | Cumulative coverage |
|---|---|---|---|---|
| 1 | Ponatinib | +18 | ABL2, DDR2, EGFR, EPHA2, EPHA4, EPHA5, EPHB2, FGFR1, FYN, MAP4K4, JAK1, LATS2, MAP3K3, PDGFRA, PDGFRB, PRKACB, TGFBR2, TEK | 18/47 (38%) |
| 2 | Sunitinib | +9 | NUAK1, AXL, CAMK2D, STK17A, MYLK, PEAK1, PRKD1, PRKD3, RPS6KA2 | 27/47 (57%) |
| 3 | Pacritinib | +5 | ACVR1, HIPK2, LIMK1, PRKCA, DAPK3 | 32/47 (68%) |
| 4 | Defactinib | +3 | BMPR2, CDK15, DYRK3 | 35/47 (74%) |
| 5 | Ceritinib | +2 | DCLK2, STK32B | 37/47 (79%) |
Per-sample EMT reversal across the Oncoscape atlas
Real tumors ranked by their mesenchymal (EMT hallmark) GSVA activation, each paired with the single drug the pathway-reversal model predicts will best reverse that tumor’s pathway state — the patient-level analog of the catalog-level coverage above.
| Tumor sample | Cancer type | EMT GSVA | Best reversal drug | Adj. score |
|---|---|---|---|---|
| SNU182_LIVER | HCC | 0.872 | Tofacitinib | 0.106 |
| JHH2_LIVER | HCC | 0.846 | Palbociclib | 0.076 |
| SNU475_LIVER | HCC | 0.743 | Palbociclib | -0.367 |
| SRR17866819 | — | 0.720 | Palbociclib | -0.096 |
| TCGA-E9-A22E-01A-11R-A157-07 | — | 0.720 | Palbociclib | -0.256 |
| DRR168609 | — | 0.710 | Tofacitinib | -0.468 |
| TCGA-BH-A6R9-01A-21R-A32P-07 | — | 0.710 | Palbociclib | -0.163 |
| 3D3C3AAA-2C7A-43C6-8137-73A6E0896FAC | — | 0.708 | Brigatinib | -0.943 |
| 20030015.LumB | — | 0.700 | Palbociclib | -0.355 |
| TCGA-AC-A23G-01A-11R-A213-07 | — | 0.700 | Palbociclib | -0.068 |
| SAMN03290911 | — | 0.700 | Abemaciclib | -0.478 |
| SRR26320076 | — | 0.700 | Palbociclib | -0.750 |
| SAMN03290935 | — | 0.700 | Palbociclib | -0.188 |
| TCGA-E2-A153-01A-12R-A12D-07 | — | 0.700 | Palbociclib | -0.117 |
| SRR25043621 | — | 0.700 | Palbociclib | -0.193 |
| SNU387_LIVER | HCC | 0.693 | Palbociclib | -0.273 |
| TCGA-AN-A0XN-01A-21R-A109-07 | — | 0.690 | Palbociclib | -0.453 |
| TCGA-AC-A23E-01A-11R-A157-07 | — | 0.690 | Palbociclib | -0.271 |
| f22cfa82-a346-4fed-bc81-21a29bf990ee | — | 0.690 | Palbociclib | -0.256 |
| 94f6e4f5-2447-4f1f-9f36-411f1c25a172 | — | 0.690 | Palbociclib | -0.204 |
Resistance-aware targeting
Which mesenchymal-program kinases are also clinical mutation drivers, and for the drugs that strongly inhibit the wild-type, whether the patient’s variant retains sensitivity or escapes (mutant residual activity from the variant assay matrix). Surfaces EMT drivers that are druggable and not escaped.
| Kinase | S8 fold-change | Variants | Cancers | WT-active drugs (escape count) |
|---|---|---|---|---|
| AXL | 5.65 | 1 | carcinoma_large_intestine, carcinoma_skin | 19 · 0 escapes |
| DDR2 | 3.68 | 2 | carcinoma_lung | 34 · 12 escapes |
| PDGFRA | 2.74 | 4 | AML-US, adnexal_tumour_skin, carcinoma_breast, carcinoma_large_intestine… | 23 · 14 escapes |
| FGFR1 | 2.73 | 2 | carcinoma_breast, carcinoma_lung, glioma_central_nervous_system | 22 · 14 escapes |
| AKT3 | 2.59 | 2 | KIRC, KIRC-US, LUSC, LUSC-US… | 1 · 0 escapes |
| FYN | 1.85 | 1 | AML-US, glioma_central_nervous_system | 15 · 0 escapes |
| ACVR1 | 1.66 | 2 | 3 · 1 escape | |
| EGFR | 1.50 | 64 | BRCA, BRCA-EU, BRCA-US, COREAD… | 24 · 23 escapes |
| MET | 1.20 | 29 | 10 · 8 escapes | |
| TEK | 1.16 | 9 | 7 · 3 escapes |
Confidence shortlist
Each of the 47 mesenchymal kinases scored on the signals we can measure: S8 mesenchymal enrichment, the best clinical inhibition achievable in S9, whether it is a mutation driver, and whether any variant escapes. Note: DepMap genetic essentiality is not yet ingested (Table S7 holds only per-lineage counts), so this is not an essentiality verdict.
| Kinase | S8 fold-change | Best residual % | Strongly druggable | Mutation driver | Resistance escape |
|---|---|---|---|---|---|
| AXL | 5.65 | 1 | ✓ | ✓ | no |
| MYLK | 4.25 | 14 | ✓ | — | — |
| PDGFRB | 3.93 | 0 | ✓ | — | — |
| DDR2 | 3.68 | 0 | ✓ | ✓ | ⚠ yes |
| PDGFRA | 2.74 | 0 | ✓ | ✓ | ⚠ yes |
| FGFR1 | 2.73 | 0 | ✓ | ✓ | ⚠ yes |
| AKT3 | 2.59 | 1 | ✓ | ✓ | no |
| TGFBR2 | 2.39 | 9 | ✓ | — | — |
| PRKCA | 2.09 | 2 | ✓ | — | — |
| CDK15 | 2.05 | 30 | ✓ | — | — |
| PRKD1 | 2.02 | 7 | ✓ | — | — |
| STK17A | 1.98 | 29 | ✓ | — | — |
| NUAK1 | 1.97 | 2 | ✓ | — | — |
| FYN | 1.85 | 0 | ✓ | ✓ | no |
| STK32B | 1.72 | 11 | ✓ | — | — |
| ABL2 | 1.66 | 0 | ✓ | — | — |
| ACVR1 | 1.66 | 3 | ✓ | ✓ | ⚠ yes |
| MAP4K4 | 1.66 | 0 | ✓ | — | — |
| RPS6KA2 | 1.66 | 1 | ✓ | — | — |
| PEAK1 | 1.59 | 0 | ✓ | — | — |
| DYRK3 | 1.57 | 0 | ✓ | — | — |
| DCLK2 | 1.52 | 22 | ✓ | — | — |
| EPHA5 | 1.50 | 0 | ✓ | — | — |
| EGFR | 1.50 | 0 | ✓ | ✓ | ⚠ yes |
| LATS2 | 1.37 | 4 | ✓ | — | — |
| PRKACB | 1.25 | 2 | ✓ | — | — |
| EPHA2 | 1.22 | 0 | ✓ | — | — |
| MET | 1.20 | 1 | ✓ | ✓ | ⚠ yes |
| CAMK2D | 1.19 | 0 | ✓ | — | — |
| JAK1 | 1.17 | 1 | ✓ | — | — |
| DAPK3 | 1.16 | 2 | ✓ | — | — |
| TEK | 1.16 | 1 | ✓ | ✓ | ⚠ yes |
| EPHB2 | 1.15 | 0 | ✓ | — | — |
| HIPK2 | 1.12 | 2 | ✓ | — | — |
| TGFBR1 | 1.12 | 15 | ✓ | — | — |
| EPHA4 | 1.10 | 0 | ✓ | — | — |
| LIMK1 | 1.10 | 1 | ✓ | — | — |
| BMPR2 | 1.10 | 16 | ✓ | — | — |
| PRKD3 | 1.04 | 8 | ✓ | — | — |
| MAP3K3 | 1.03 | 6 | ✓ | — | — |
| SGK1 | 2.88 | 42 | — | — | — |
| PLK2 | 2.42 | 45 | — | — | — |
| CAMK1 | 1.86 | 37 | — | — | — |
| NEK6 | 1.67 | 89 | — | — | — |
| NEK7 | 1.54 | 84 | — | — | — |
| MAP3K12 | 1.29 | 37 | — | — | — |
| CAMK4 | 1.06 | 76 | — | — | — |