Comprehensive exploration of graphically defined reaction spaces

Abstract

Existing reaction transition state (TS) databases are comparatively small and lack chemical diversity. Here, this data gap has been addressed using the concept of a graphically-defined model reaction to comprehensively characterize a reaction space associated with C, H, O, and N containing molecules with up to 10 heavy (non-hydrogen) atoms. The resulting dataset is composed of 176,992 organic reactions possessing at least one validated TS, activation energy, heat of reaction, reactant and product geometries, frequencies, and atom-mapping. For 33,032 reactions, more than one TS was discovered by conformational sampling, allowing conformational errors in TS prediction to be assessed. Data is supplied at the GFN2-xTB and B3LYP-D3/TZVP levels of theory. A subset of reactions were recalculated at the CCSD(T)-F12/cc-pVDZ-F12 and ωB97X-D2/def2-TZVP levels to establish relative errors. The resulting collection of reactions and properties are called the Reaction Graph Depth 1 (RGD1) dataset. RGD1 represents the largest and most chemically diverse TS dataset published to date and should find immediate use in developing novel machine learning models for predicting reaction properties.

Graph to Activation Energy Models Easily Reach Irreducible Errors but Show Limited Transferability

“…The Chemprop and EGAT models used in this paper were trained on the RGD1 data set, which contains around 177,000 reactions with up to 10 heavy atoms consisting of carbon, hydrogen, nitrogen, and oxygen. In brief, the RGD1 data set was generated by a graph-based enumeration of ∼700,000 reactions involving reactants sampled from PubChem. , A reaction conformational sampling strategy , was applied to generate up to three conformations for each reaction that were used to initialize double-ended TS searches, followed by Berny optimization, and intrinsic reaction coordinate (IRC) validation at the GFN2-xTB level of theory.…”

Atomic fragment approximation from a tensor network

“…The reaction graph depth (RGD1) dataset 33 is implemented to test the performance of chemical reaction prediction. It contains 176 992 organic reactions with validated transition states, activation energy, heat of reaction, reactant and product geometries, frequencies, and atom mapping.…”

Accelerating Reaction Network Explorations with Automated Reaction Template Extraction and Application

“…We would like to note that it would also be possible to describe this approach as tracking additional information about the columns and rows in the so-called bond and electron (BE) matrix . The counting of elements in such a BE matrix is similar to how some automated reaction exploration software packages restrict explorations. − …”