Reproduce “2DNMRGym: An Annotated Experimental Dataset for Atom-Level Molecular Representation Learning in 2D NMR via Surrogate Supervision” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Yunrui Li, Hao Xu, Pengyu Hong
2DNMRGym creates a dataset linking experimental HSQC peaks to molecular graphs and atom assignments, using algorithm-generated labels for training and expert annotations for evaluation. The paper reports 21,869 algorithm-labeled and 479 expert-labeled spectra in its introduction, but 22,157 total later. On held-out expert data, GIN-Transformer has lower carbon and hydrogen MAEs than ComENet; the task is assigned HSQC C–H shift prediction, not structure elucidation.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.