Tc-43/OX2R_Agonist_Designs_7SQO

Orexin-2 Receptor (OX2R) Agonist Designs — 7SQO / GA-II

Why this target matters. Orexin-2 agonists are the first mechanism to address the cause of narcolepsy type 1 — loss of orexin-producing neurons — rather than managing its symptoms with stimulants.

878 de novo small-molecule designs docked into the orthosteric agonist pocket of the human orexin-2 receptor, using the active-state OX2R–G_i cryo-EM structure 7SQO (OX2R bound to TAK-925, G_i and scFv16, 3.17 Å) as the receptor.

Orexin-2 agonists are of interest for narcolepsy type 1 and other disorders of hypersomnolence, where orexin neuron loss is the underlying pathology.

Contents

File	Description
designs.csv	Index of all 878 designs: ID, canonical SMILES, docking score, physicochemical properties, per-design contact residues
ligands.sdf	All 878 docked poses as a single 3D SDF, bond orders and formal charges assigned, stereo perceived from 3D
structures/	Per-design ligand-only PDB (<ligand_id>.pdb), 878 files, in the receptor frame
receptor_7SQO.pdb	The single shared receptor used for all docking runs

The receptor is byte-identical across every docking run, so it is stored once. Ligand poses are in the receptor's coordinate frame — load receptor_7SQO.pdb together with any ligand file or SDF record with no superposition needed.

Receptor chains: R orexin receptor type 2 (residues 55–354, plus 4 oleic acid molecules modelled in the original structure), A G_iα1, B Gβ1, G Gγ2, E scFv16. All designed ligands contact chain R only.

designs.csv columns

ligand_id, smiles, vina_score (kcal/mol), mw, clogp, tpsa, hbd, hba, rotb, rings, heavy_atoms, formal_charge, qed, contact_residues (receptor residues within 4 Å of the pose, ;-separated).

Rows are sorted by docking score, best first.

Set profile

	min	median	max
AutoDock Vina score (kcal/mol)	−13.7	−11.5	−11.0
Molecular weight	365.4	428.4	628.3
cLogP	−4.2	1.9	5.7
TPSA (Ų)	58.6	111.5	200.9
QED	0.20	0.50	0.80
Heavy atoms	27	31	33

810 of the 878 SMILES are unique (the remainder are distinct poses or stereoisomers of a shared 2D structure). 837/878 (95%) satisfy Lipinski's rule of five; 387 carry a net formal charge at the modelled protonation state.

Binding site

Residues contacted by ≥50% of the designs, with frequency:

Q134 98% · H350 98% · N324 94% · V114 94% · T135 93% · P131 90% · I320 87% · K327 77% · W120 72% · R328 72% · V138 71% · T111 65% · Q187 64% · F227 60% · F346 58% · C210 56%

This recapitulates the OX2R orthosteric pocket — Q134^3.32 and H350^7.39, the anchors engaged by TAK-925 and by the dual antagonist suvorexant, are contacted by essentially every design. Median 17 contact residues per design.

Usage

from datasets import load_dataset

ds = load_dataset("Tc-43/OX2R_Agonist_Designs_7SQO")["train"]
top = ds.sort("vina_score")[:10]

# 3D poses
from rdkit import Chem
mols = [m for m in Chem.SDMolSupplier("ligands.sdf") if m is not None]
mols[0].GetProp("vina_score")

# single design in its receptor
pymol receptor_7SQO.pdb structures/43_7SQO_II260802205721795319358a.pdb

Caveats

Docking scores are AutoDock Vina estimates from a single rigid-receptor run against one cryo-EM conformation. They rank poses, not affinities, and the dynamic range here is narrow (−13.7 to −11.0) because the set is already score-filtered. Nothing in this set has been synthesised or assayed — no compound has measured OX2R binding, functional agonism, selectivity over OX1R, or ADME/tox data. Treat these as computational starting points for triage and medicinal-chemistry review, not as validated agonists.

Poses are rigid-receptor and do not model induced fit; protonation states are as assigned by the docking preparation and may not reflect the bound state.

Citation

@misc{orexin_agonist_designs_7sqo,
  title  = {Orexin-2 Receptor Agonist Designs (7SQO / GA-II)},
  author = {TechnetiumTx},
  year   = {2026},
  url    = {https://huggingface.co/datasets/Tc-43/OX2R_Agonist_Designs_7SQO}
}

Receptor structure: PDB 7SQO — Structure of the orexin-2 receptor (OX2R) bound to TAK-925, Gi and scFv16.