English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

A proof-of-concept framework for identifying molecules of unknown elementalcomposition and structure using experimental rotational data and probabilisticdeep learning is presented. Using a minimal set of input data determinedexperimentally, we describe four neural network architectures that yieldinformation to assist in the identification of an unknown molecule. The firstarchitecture translates spectroscopic parameters into Coulomb matrixeigenspectra, as a method of recovering chemical and structural informationencoded in the rotational spectrum. The eigenspectrum is subsequently used bythree deep learning networks to constrain the range of stoichiometries,generate SMILES strings, and predict the most likely functional groups presentin the molecule. In each model, we utilize dropout layers as an approximationto Bayesian sampling, which subsequently generates probabilistic predictionsfrom otherwise deterministic models. These models are trained on a modestlysized theoretical dataset comprising ${\sim}$83,000 unique organic molecules(between 18 and 180 amu) optimized at the $\omega$B97X-D/6-31+G(d) level oftheory where the theoretical uncertainty of the spectroscopic constants arewell understood and used to further augment training. Since chemical andstructural properties depend highly on molecular composition, we divided thedataset into four groups corresponding to pure hydrocarbons, oxygen-bearing,nitrogen-bearing, and both oxygen- and nitrogen-bearing species, training eachtype of network with one of these categories thus creating "experts" withineach domain of molecules. We demonstrate how these models can then be used forpractical inference on four molecules, and discuss both the strengths andshortcomings of our approach, and the future directions these architectures cantake.

Molecule Identification with Rotational Spectroscopy and Probabilistic Deep Learning