End-to-end molecular structure elucidation from multimodal NMR spectra images using vision transformers

Chao Han, Xiaolin Pan, Yingkai Zhang · 2026

Read End-to-end molecular structure elucidation from multimodal NMR spectra images using vision transformers by Chao Han, Xiaolin Pan, Yingkai Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/pmc-13211313.txt.

Description

Reading a molecule's structure from its magnetic spectra is expert work, and a 2026 study by Chao Han automates it with vision transformers. The research article introduces NMRViT, which takes raw one- and two-dimensional NMR spectrum images directly, instead of hand-picked peak lists, and predicts the molecular structure end to end.

Training ran on a large simulated spectral library before testing on real experimental benchmarks. Moving from simulation to the laboratory left a clear performance gap, but fine-tuning on a handful of real spectra shrank it substantially in the authors' experiments. A final re-ranking step, scoring candidates against expected chemical shifts, lifted accuracy in both settings.

Catalog record

CREATORChao Han, Xiaolin Pan, Yingkai Zhang
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE45 KB
VERSION1.0
LICENSECC BY 4.0
RIGHTSCC BY
SOURCEEurope PMC

Source: https://europepmc.org/articles/PMC13211313

Browse the full library · About the Tower of Babel