Skip to main content
AACR Project GENIE v19.0 · 21,017 myeloid samples Panel-adjusted Fisher's exact with Benjamini-Hochberg FDR N=1 case study · Not clinical guidance

Molecular Docking & Drug Pipeline

SHP2 inhibitors vs PTPN11 E76Q, IDH2 inhibitors vs IDH2 R140Q · Boltz-2 affinity, DiffDock 2.0

Best PTPN11 Binder
RMC-4550
Boltz-2 pIC50 6.21 · DiffDock 2.0 -1.65 i
Best IDH2 Binder
Enasidenib
Boltz-2 pIC50 5.23 · FDA 2017 i
DiffDock 2.0 Top (SHP2)
SHP099
Confidence -1.53 · top of 10 poses · low band (below -1.5) i
Boltz-1 Runs Ligand invalid
4/4
Vast.ai A100 · wrong SMILES, superseded 15 Apr 2026

Binding Affinity Comparison

Boltz-2 predicted pIC50 (6 minus affinity_pred_value; NVIDIA NIM, 15 Apr 2026), ligands from PubChem-verified SMILES. Higher pIC50 means tighter predicted binding; pIC50 6 corresponds to 1 µM. AG-221 is enasidenib run a second time (same PubChem CID), so the gap between those two bars is run-to-run noise. Hover a bar for the Boltz-2 binding probability (below 50% for all three SHP2 inhibitors) and the DiffDock 2.0 top-pose confidence 33
[33] G 2023
DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking. ICLR (2023)
for the same pair. Ranking for PTPN11 E76Q: Boltz-2 RMC-4550 > TNO155 > SHP099; DiffDock 2.0 SHP099 > RMC-4550 > TNO155 (all three in DiffDock's low-confidence band, below -1.5).

Superseded Docking Runs

The AutoDock Vina, DiffDock v1 and Boltz-1 results below were computed against the wrong molecules. On 15 Apr 2026 a QA pass checked every drug SMILES string against PubChem and found that all of them were syntactically valid but encoded different compounds (TNO155 off by +70.6 Da, RMC-4550 +47.6 Da, ivosidenib −236.3 Da). The scores are kept here so earlier reports can be reconciled, and are not evidence for any drug. The reruns above use the corrected SMILES. The methods do not agree on the PTPN11 E76Q ranking: Boltz-2 puts RMC-4550 first, DiffDock 2.0 puts SHP099 first, and the invalid Vina run put TNO155 first.
Target Drug Vina (kcal/mol) DiffDock v1 Status
PTPN11 E76Q RMC-4550 -8.37 -9.26 Invalid ligand
PTPN11 E76Q SHP099 -7.79 -7.83 Invalid ligand
PTPN11 E76Q TNO155 -9.34 -7.49 Invalid ligand
IDH2 R140Q Enasidenib -9.43 Not run Invalid ligand
IDH2 R140Q Ivosidenib -10.36 Not run Invalid ligand
AutoDock Vina 1.2.x (local, RTX 4060) 32
[32] O 2010
AutoDock Vina: improving the speed and accuracy of docking with a new scoring function, efficient optimization, and multithreading. J Comput Chem (2010)
and DiffDock v1 NIM 33
[33] G 2023
DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking. ICLR (2023)
, March 2026, with hallucinated ligand SMILES. Struck through on purpose.

Drug Pipeline

Drug Target Mechanism Phase Best Score Evidence Source Notes
Enasidenib 21
[21] EM 2017
Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood (2017)
IDH2 R140Q IDH2 inhibitor FDA Approved (2017) pIC50 5.23 · DD2 -1.40 OncoKB Level 1, PharmGKB Testing Required Standard for IDH2-mutant AML
Vorasidenib IDH2 Dual IDH1/2 inhibitor FDA (glioma 2024) / Phase I (AML) pIC50 4.76 · DD2 -0.82 DGIdb, ClinicalTrials.gov AML: Phase I only (AG-881, NCT02492737, completed)
TNO155 23
[23] MJ 2020
Identification of TNO155, an Allosteric SHP2 Inhibitor for the Treatment of Cancer. J Med Chem (2020)
PTPN11 E76Q SHP2 allosteric inhibitor No hematological trial (Phase I/II in solid tumours) pIC50 5.41 · DD2 -1.68 DGIdb confirmed (as batoprotafib) Novartis lead compound
RMC-4550 PTPN11 E76Q SHP2 allosteric inhibitor Preclinical pIC50 6.21 · DD2 -1.65 Literature (FLT3/KIT-mutant AML models) 22
[22] B 2023
Allosteric SHP2 inhibition increases apoptotic dependency on BCL2 and synergizes with venetoclax in FLT3- and KIT-mutant AML. Cell Rep Med (2023)
Revolution Medicines
Trametinib MEK (PTPN11→MAPK) MEK inhibitor FDA (melanoma) Not docked DGIdb Convergence node for PTPN11+SETBP1
SHP099 PTPN11 E76Q SHP2 allosteric inhibitor Tool compound pIC50 5.13 · DD2 -1.53 Literature Preclinical only
Venetoclax BCL-2 BH3 mimetic FDA Approved (2018) Not docked PharmGKB Backbone with azacitidine
Azacitidine DNMT3A Hypomethylating agent FDA Approved Not docked PharmGKB, DGIdb Standard backbone
FTY720 1
[1] R 2013
Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet (2013)
PP2A (SETBP1) PP2A activator Preclinical (AML) Not docked Literature MS-approved drug repurposing
Best score: Boltz-2 predicted pIC50 and DiffDock 2.0 top-pose confidence (DD2: above 0 high, -1.5 to 0 moderate, below -1.5 low), both with PubChem-verified SMILES (15 Apr 2026). Phase refers to highest clinical development stage for the relevant hematological indication. Evidence sources: DGIdb 38
[38] SL 2021
Integration of the Drug-Gene Interaction Database (DGIdb 4.0) with open crowdsource efforts. Nucleic Acids Res (2021)
, PharmGKB 39
[39] M 2021
An Evidence-Based Framework for Evaluating Pharmacogenomics Knowledge for Personalized Medicine. Clin Pharmacol Ther (2021)
, OncoKB 35
[35] D 2017
OncoKB: A Precision Oncology Knowledge Base. JCO Precis Oncol (2017)
. The DepMap 46
[46] A 2017
Defining a Cancer Dependency Map. Cell (2017)
query returned no data.

Drug Sensitivity (Literature)

The DepMap query for these genes and compounds returned no data, so no DepMap analysis was done. Published cell-line evidence: in FLT3- and KIT-mutant AML cell lines, the SHP2 inhibitor RMC-4550 lowered MCL1, raised BMF and increased apoptotic dependency on BCL2, and RMC-4550 plus venetoclax was synergistically lethal in cell lines and xenografts 22
[22] B 2023
Allosteric SHP2 inhibition increases apoptotic dependency on BCL2 and synergizes with venetoclax in FLT3- and KIT-mutant AML. Cell Rep Med (2023)
. PTPN11-mutant cells were not part of that study.

Whether SHP2 inhibition would add to venetoclax in a PTPN11 E76Q clone is therefore an extrapolation from RTK-driven AML, not a tested result.

Synthetic Lethality

Gene Pair Interaction Type Evidence Level Therapeutic Implication
DNMT3A + EZH2 Functional association (STRING) STRING 0.999 44
[44] D 2023
The STRING database in 2023: protein-protein association networks and functional enrichment analyses for any sequenced genome of interest. Nucleic Acids Res (2023)
Highest STRING score of the 10 patient-gene pairs, mostly from text mining (0.998)
IDH2 + SETBP1 Possible synthetic lethality (hypothesis) Literature-curated; no SynLethDB hit Not mutually exclusive in GENIE: 20 samples observed vs 22.0 expected (O/E 0.91)
PTPN11 + BCL2 Cooperative (SHP2 + BCL2 inhibition) Literature (FLT3/KIT-mutant AML) 22
[22] B 2023
Allosteric SHP2 inhibition increases apoptotic dependency on BCL2 and synergizes with venetoclax in FLT3- and KIT-mutant AML. Cell Rep Med (2023)
SHP2i + venetoclax synergy shown in RTK-driven AML, not in PTPN11-mutant cells
EZH2 + Mono7 Haploinsufficiency Literature 10
[10] A 2020
Mutational mechanisms of EZH2 inactivation in myeloid neoplasms. Leukemia (2020)
9
[9] T 2010
Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders. Nat Genet (2010)
EZH2 LoF + chr7 TSG loss compound
Gene pair interactions relevant to this mutation profile. Evidence from STRING, a literature-curated synthetic lethality list (the SynLethDB API returned no usable data), GENIE co-occurrence and published literature.

Boltz-1 Co-folding

Ligands in these four runs were wrong. The same hallucinated SMILES strings as in the superseded Vina and DiffDock v1 runs were used, so the binding geometries described here are not valid evidence (found 15 Apr 2026). Protein-only conclusions stand; the Boltz-2 affinities above are the valid replacement.

4 drug-protein complexes completed via Boltz-1 34
[34] J 2024
Boltz-1: Democratizing Biomolecular Interaction Modeling. bioRxiv (2024)
on Vast.ai A100:
  • IDH2 R140Q + Enasidenib 21
    [21] EM 2017
    Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood (2017)
    : ligand MW 490.5 Da, PubChem enasidenib 473.4 Da
  • IDH2 R140Q + AG-221 21
    [21] EM 2017
    Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood (2017)
    : same wrong ligand as the enasidenib run (AG-221 is enasidenib)
  • PTPN11 E76Q + TNO155 23
    [23] MJ 2020
    Identification of TNO155, an Allosteric SHP2 Inhibitor for the Treatment of Cancer. J Med Chem (2020)
    : ligand MW 492.5 Da, PubChem TNO155 421.9 Da
  • PTPN11 E76Q + RMC-4550: ligand MW 485.0 Da, PubChem RMC-4550 437.4 Da
At the time, all four predictions appeared to confirm the binding geometries from the AutoDock Vina and DiffDock v1 analyses. All three methods used the same invalid ligands, so that agreement is not evidence that these drugs engage their targets in the mutant context.

Coverage gaps: DNMT3A R882H is not directly druggable (addressed by hypomethylating agents). SETBP1 G870S 1
[1] R 2013
Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet (2013)
has no known small-molecule inhibitors; G870S raises SET and lowers PP2A activity 1
[1] R 2013
Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet (2013)
, so PP2A activation (FTY720/fingolimod) is the closest experimental strategy.

Treatment Strategy

Recommended backbone: venetoclax + azacitidine + enasidenib 21
[21] EM 2017
Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood (2017)

This combination addresses DNMT3A R882H (via hypomethylation) and the small IDH2 R140Q subclone, VAF 2% (via direct inhibition). Enasidenib 21
[21] EM 2017
Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood (2017)
is FDA-approved for relapsed/refractory IDH2-mutant AML. As a triplet it has been given with venetoclax and oral decitabine, not azacitidine (DiNardo et al., J Clin Oncol 2025, PMID 40513054: 60 IDH1/IDH2-mutant patients across arms, CRc 92%). Risk classification: ELN 2022 Adverse 14
[14] H 2022
Diagnosis and management of AML in adults: 2022 recommendations from an international expert panel on behalf of the ELN. Blood (2022)
.

SHP2 inhibitors (TNO155 23
[23] MJ 2020
Identification of TNO155, an Allosteric SHP2 Inhibitor for the Treatment of Cancer. J Med Chem (2020)
, RMC-4550):
no myeloid trial. TNO155 trials are in solid tumours only, RMC-4550 is preclinical, and the one AML study with an allosteric SHP2 inhibitor (ERAS-601, NCT05279859) was withdrawn. PTPN11 E76Q activates the RAS-MAPK pathway; allosteric SHP2 inhibition locks the phosphatase in its autoinhibited state. These agents represent the most promising targeted option for the PTPN11 subclone. In FLT3- and KIT-mutant AML cell lines, SHP2 inhibition lowered MCL1 and synergised with venetoclax 22
[22] B 2023
Allosteric SHP2 inhibition increases apoptotic dependency on BCL2 and synergizes with venetoclax in FLT3- and KIT-mutant AML. Cell Rep Med (2023)
; PTPN11-mutant cells were not tested.

No direct SETBP1 inhibitors exist. SETBP1 G870S disrupts the SKI domain degron motif 1
[1] R 2013
Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet (2013)
, stabilizing the protein and suppressing PP2A tumor suppressor activity via higher SET levels 1
[1] R 2013
Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet (2013)
. FTY720 (fingolimod), approved for multiple sclerosis, reactivates PP2A and has preclinical activity in AML cell models; no study in SETBP1-mutant AML was found in PubMed. This represents an experimental drug repurposing strategy.
CRITICAL: Tazemetostat (EZH2 inhibitor) is CONTRAINDICATED for this patient. EZH2 V662A is predicted loss-of-function: SET-domain missense mutations abrogate catalytic activity 10
[10] A 2020
Mutational mechanisms of EZH2 inactivation in myeloid neoplasms. Leukemia (2020)
9
[9] T 2010
Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders. Nat Genet (2010)
, but V662A itself has not been tested. Tazemetostat inhibits EZH2 activity that is already lost. Administering an EZH2 inhibitor would provide no therapeutic benefit and could exacerbate the epigenetic dysregulation caused by the existing EZH2 LoF mutation.
References
  1. Stein EM et al. Enasidenib in mutant IDH2 relapsed or refractory acute myeloid leukemia. Blood (2017). PubMed
  2. LaMarche MJ et al. Identification of TNO155, an Allosteric SHP2 Inhibitor for the Treatment of Cancer. J Med Chem (2020). DOI
  3. Popescu B et al. Allosteric SHP2 inhibition increases apoptotic dependency on BCL2 and synergizes with venetoclax in AML. Cell Rep Med (2023). DOI
  4. Piazza R et al. Recurrent SETBP1 mutations in atypical chronic myeloid leukemia. Nat Genet (2013). PubMed
  5. Piazza R et al. SETBP1 induces transcription of a network of development genes by acting as an epigenetic hub. Nat Commun (2018). DOI
  6. Chase A et al. Mutational mechanisms of EZH2 inactivation in myeloid neoplasms. Leukemia (2020). DOI
  7. Ernst T et al. Inactivating mutations of the histone methyltransferase gene EZH2 in myeloid disorders. Nat Genet (2010). PubMed
  8. Döhner H et al. Diagnosis and management of AML in adults: 2022 ELN recommendations. Blood (2022). PubMed
  9. Trott O, Olson AJ. AutoDock Vina: improving the speed and accuracy of docking. J Comput Chem (2010). DOI
  10. Corso G et al. DiffDock: Diffusion Steps, Twists, and Turns for Molecular Docking. ICLR (2023). DOI
  11. Passaro S et al. Boltz-2: Towards Accurate and Efficient Binding Affinity Prediction. bioRxiv (2025). DOI
  12. Wohlwend J et al. Boltz-1: Democratizing Biomolecular Interaction Modeling. bioRxiv (2024). DOI