Determining the chemical composition of unknown compounds plays a central role in chemical analysis. Searching mass spectra against experimental or in-silico spectra libraries are the most widely used molecular identification methods. With the high complexity of mass spectra in mind, we aim to make completely de novo predictions of small molecule structures corresponding to mass spectra. We build up a promising methodology by minimizing the complexity of electron-ionization mass spectrum, and show it reveals molecular structures with high accuracy. Though our novel approach is capable of producing possible candidates directly via machine translation, it also complements conventional methods with newly developed scoring strategy.

Reducing complexity of mass spectra : Atom-type approach for molecular identification

Fall 2022 abstracts, posters, and presentations will launch in mid-September.

ACS Fall 2022

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.