MolDiscovery: learning mass spectrometry fragmentation of small molecules

Abstract

Identification of small molecules is a critical task in various areas of life science. Recent advances in mass spectrometry have enabled the collection of tandem mass spectra of small molecules from hundreds of thousands of environments. To identify which molecules are present in a sample, one can search mass spectra collected from the sample against millions of molecular structures in small molecule databases. The existing approaches are based on chemistry domain knowledge, and they fail to explain many of the peaks in mass spectra of small molecules. Here, we present molDiscovery, a mass spectral database search method that improves both efficiency and accuracy of small molecule identification by learning a probabilistic model to match small molecules with their mass spectra. A search of over 8 million spectra from the Global Natural Product Social molecular networking infrastructure shows that molDiscovery correctly identify six times more unique small molecules than previous methods.

Cross-Modal Retrieval between 13C NMR Spectra and Structures for Compound Identification Using Deep Contrastive Learning

“…To analyze the impact of the various compound classes on the performance of CReSS, we classified the compounds in the external test set into 15 superclasses defined by ClassyFire. , As shown in Figure S8, CReSS performed well in all 15 categories. Recall@10 maintained at least 60% and were higher than 80% in 11 of the 15 categories.…”

“…Library matching is a vital and powerful compound identification technique that has been commonly utilized in identifying molecular samples by comparing the query spectrum with each entry in the reference spectral library. − Carbon-13 nuclear magnetic resonance ( 13 C NMR) spectroscopy has received considerable attention because of its power as a probe of the skeletal backbone of an organic compound, whose information is not as readily accessible by other spectroscopic methods . Due to a large chemical shift range and narrow peak width, a 13 C NMR spectrum can be easily converted to a simple numerical list of signal positions with minimal loss of information, which contributes to the rising and wide use of 13 C spectral library matching systems in compound identification systems .…”

SIMILE enables alignment of tandem mass spectra with statistical significance

“…The fragmentation process in tandem mass spectrometry can be modeled as a network of fragmentation paths connecting precursor ions to product ions with fragmentation reactions 34 . In addition, it is broadly assumed that similar structures fragment similarly.…”

Metabolomics Approaches to Dereplicate Natural Products from Coral-Derived Bioactive Bacteria

“…Additional compound annotations were performed by searching for entries in databases such as The Natural Product Atlas and MarinLit and by the use of in silico tools such as MolDiscovery (v.1.0.0), SIRIUS 4.9.3 with CSI:FingerID, and CANOPUS and through literature searches for compounds known to be produced by marine bacteria for untargeted analysis. MolDiscovery, available through the GNPS platform, compares in silico-generated mass spectra of small molecules from a variety of databases with experimental MS 2 spectra and includes a similarity score for each reported match . The .mgf file generated from Mzmine on the positive mode data was submitted to the MolDiscovery workflow.…”

“…MolDiscovery, available through the GNPS platform, compares in silico-generated mass spectra of small molecules from a variety of databases with experimental MS 2 spectra and includes a similarity score for each reported match. 76 The .mgf file generated from Mzmine on the positive mode data was submitted to the MolDiscovery workflow. Proposed annotations were further supported either through a comparison with published MS 2 spectra or through analytical standards.…”