FOdorant conformational ensembles encode topology, molecular size and polar-surface coherence, not shape

Odor is widely modelled as three-dimensional shape recognition: a flexible odorant presents a 300 K conformer ensemble to a receptor repertoire, implying that perceptual breadth peaks at an intermediate number of accessible shapes and that explicit 3D representations outperform 2D molecular graphs. We test both predictions under a cryptographically pre-registered plan across catalogue profiles (n = 4,808), human psychophysics (n = 304) and heterologous receptor arrays (n = 294). Computing 300 K GFN2-xTB ensembles at a 200-conformer ceiling for all 1,821 flexible cohort members (106,526 conformers) and benchmarking 21 neural representations across 10 seeds, we find that neither claim holds. Conformational entropy saturates by 28 conformers (88% attained at a ceiling of 20), excluding an interior breadth optimum sixfold below the populated range (β2 = +0.0080, 95% CI -0.0014 to +0.0172). No 3D representation outperforms a capacity-matched or descriptor-augmented 2D graph. Out-of-sample-replicated confound analyses show that 3D shape dispersion vanishes once chain topology is controlled; instead, breadth reflects coarse topological flexibility (peaking at 9.2 rotatable bonds), molecular extent and polar-surface coherence (concave β2 = -0.0264). Keywords: conformational ensemble · odor breadth · molecular representation · Artificial Intelligence · Graph Neural Network

Authors

Institutions

Publication Details

Journal
ChemRxiv
Published
2026-11-18
DOI
https://doi.org/10.26434/chemrxiv.15012459/v1
Primary Topic
Computational Drug Discovery Methods
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

FOdorant conformational ensembles encode topology, molecular size and polar-surface coherence, not shape

Yuan Shang
ChemRxiv
Computational Drug Discovery Methods
preprint

FOdorant conformational ensembles encode topology, molecular size and polar-surface coherence, not shape

Yuan Shang
preprint en

Abstract

Odor is widely modelled as three-dimensional shape recognition: a flexible odorant presents a 300 K conformer ensemble to a receptor repertoire, implying that perceptual breadth peaks at an intermediate number of accessible shapes and that explicit 3D representations outperform 2D molecular graphs. We test both predictions under a cryptographically pre-registered plan across catalogue profiles (n = 4,808), human psychophysics (n = 304) and heterologous receptor arrays (n = 294). Computing 300 K GFN2-xTB ensembles at a 200-conformer ceiling for all 1,821 flexible cohort members (106,526 conformers) and benchmarking 21 neural representations across 10 seeds, we find that neither claim holds. Conformational entropy saturates by 28 conformers (88% attained at a ceiling of 20), excluding an interior breadth optimum sixfold below the populated range (β2 = +0.0080, 95% CI -0.0014 to +0.0172). No 3D representation outperforms a capacity-matched or descriptor-augmented 2D graph. Out-of-sample-replicated confound analyses show that 3D shape dispersion vanishes once chain topology is controlled; instead, breadth reflects coarse topological flexibility (peaking at 9.2 rotatable bonds), molecular extent and polar-surface coherence (concave β2 = -0.0264). Keywords: conformational ensemble · odor breadth · molecular representation · Artificial Intelligence · Graph Neural Network

ChemRxiv
Zhengzhou University (CN)
Computational Drug Discovery Methods
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon