Quantum Chemistry Calculations for Metabolomics

Borges, Ricardo Moreira; Colby, Sean; Das, Susanta; Edison, Arthur S.; Fiehn, Oliver; Kind, Tobias; Lee, Jesi; Merrill, Amy T.; Merz, Kenneth M.; Metz, Thomas O.; Nuñez, Jamie; Tantillo, Dean J.; Wang, Lee‐Ping; Renslow, Ryan

doi:10.1021/acs.chemrev.0c00901

Cited by 60 publications

(49 citation statements)

References 535 publications

(984 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In silico prediction methods for NMR and MS/MS have improved accuracy although ambiguity is expected to remain for large molecular weight formulas 67 . 1 H and 13 C 1D NMR and MS/MS fragmentation in silico predictions can be prioritized for target features identified with this approach 66 .…”

Section: Discussionmentioning

confidence: 99%

An anchored experimental design and meta-analysis approach to address batch effects in large-scale metabolomics

Shaver¹,

Garcia²,

Gouveia³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Large-scale untargeted metabolomics studies suffer from individual variation, batch effects and instrument variability, making comparisons of common spectral features across studies difficult. One solution is to compare studies after compound identification. However, compound identification is expensive and time consuming. We successfully identify common spectral features across multiple studies, with a generalizable experimental design approach. First, we included an anchor strain, PD1074, during sample and data collection. Second, we collected data in blocks with multiple controls. These anchors enabled us to successfully integrate three studies of Caenorhabditis elegans for nuclear magnetic resonance (NMR) spectroscopy and liquid chromatography-mass spectrometry (LC-MS) data from five different assays. We found 34% and 14% of features to be significant in LC-MS and NMR, respectively. Between 20-50% of spectral features differ in a mutant and among a set of genetically diverse natural strains, suggesting this reduced set of spectral features are excellent targets for compound identification.

show abstract

Section: Discussionmentioning

confidence: 99%

An anchored experimental design and meta-analysis approach to address batch effects in large-scale metabolomics

Shaver¹,

Garcia²,

Gouveia³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Consequently, the likelihood that any molecule detected in a metabolomics experiment is actually in a library (or even close to a molecule that is [ 10 ]) is quite small, and most are not [ 11 , 12 , 13 , 14 , 15 ]. Experimentally, commonly just 10% of molecules can be identified from what are reproducible spectral features in complex matrices such as serum (e.g., [ 16 , 17 , 18 , 19 , 20 ]), despite the existence of many heuristics such as neutral mass loss, isotope patterns, and so on [ 21 , 22 , 23 , 24 ]. Solving the mass spectral molecular identification problem is thus seen widely as the key unsolved problem of metabolomics [ 4 , 11 , 14 , 15 , 19 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 ].…”

Section: Introductionmentioning

confidence: 99%

“…Experimentally, commonly just 10% of molecules can be identified from what are reproducible spectral features in complex matrices such as serum (e.g., [ 16 , 17 , 18 , 19 , 20 ]), despite the existence of many heuristics such as neutral mass loss, isotope patterns, and so on [ 21 , 22 , 23 , 24 ]. Solving the mass spectral molecular identification problem is thus seen widely as the key unsolved problem of metabolomics [ 4 , 11 , 14 , 15 , 19 , 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 , 33 , 34 , 35 , 36 , 37 , 38 , 39 , 40 ]. It was also seen as a classical problem in the early development of ‘artificial intelligence’ [ 41 , 42 , 43 , 44 ].…”

Section: Introductionmentioning

confidence: 99%

“…By contrast to the ‘inverse problem’ that we are seeking to solve (where we have a mass spectrum and seek to find the molecule that generated it), the ‘forward problem’ of mass spectrometry is considerably more tractable (especially for high-energy electron impact mass spectra [ 61 ]). Given a known chemical structure, it is possible to fragment the weakest bonds in silico and thereby generate a reasonably accurate ‘mass spectrum’ of the fragments so generated (e.g., [ 19 , 62 , 63 , 64 , 65 , 66 , 67 , 68 , 69 , 70 , 71 ]). Importantly, the advent of high-mass-resolution spectra means that each mass or fragment can be contributed by only a relatively small number of (biologically) feasible [ 72 , 73 ] molecular formulae.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

MassGenie: A Transformer-Based Deep Learning Method for Identifying Small Molecules from Their Mass Spectra

et al. 2021

View full text Add to dashboard Cite

The ‘inverse problem’ of mass spectrometric molecular identification (‘given a mass spectrum, calculate/predict the 2D structure of the molecule whence it came’) is largely unsolved, and is especially acute in metabolomics where many small molecules remain unidentified. This is largely because the number of experimentally available electrospray mass spectra of small molecules is quite limited. However, the forward problem (‘calculate a small molecule’s likely fragmentation and hence at least some of its mass spectrum from its structure alone’) is much more tractable, because the strengths of different chemical bonds are roughly known. This kind of molecular identification problem may be cast as a language translation problem in which the source language is a list of high-resolution mass spectral peaks and the ‘translation’ a representation (for instance in SMILES) of the molecule. It is thus suitable for attack using the deep neural networks known as transformers. We here present MassGenie, a method that uses a transformer-based deep neural network, trained on ~6 million chemical structures with augmented SMILES encoding and their paired molecular fragments as generated in silico, explicitly including the protonated molecular ion. This architecture (containing some 400 million elements) is used to predict the structure of a molecule from the various fragments that may be expected to be observed when some of its bonds are broken. Despite being given essentially no detailed nor explicit rules about molecular fragmentation methods, isotope patterns, rearrangements, neutral losses, and the like, MassGenie learns the effective properties of the mass spectral fragment and valency space, and can generate candidate molecular structures that are very close or identical to those of the ‘true’ molecules. We also use VAE-Sim, a previously published variational autoencoder, to generate candidate molecules that are ‘similar’ to the top hit. In addition to using the ‘top hits’ directly, we can produce a rank order of these by ‘round-tripping’ candidate molecules and comparing them with the true molecules, where known. As a proof of principle, we confine ourselves to positive electrospray mass spectra from molecules with a molecular mass of 500Da or lower, including those in the last CASMI challenge (for which the results are known), getting 49/93 (53%) precisely correct. The transformer method, applied here for the first time to mass spectral interpretation, works extremely effectively both for mass spectra generated in silico and on experimentally obtained mass spectra from pure compounds. It seems to act as a Las Vegas algorithm, in that it either gives the correct answer or simply states that it cannot find one. The ability to create and to ‘learn’ millions of fragmentation patterns in silico, and therefrom generate candidate structures (that do not have to be in existing libraries) directly, thus opens up entirely the field of de novo small molecule structure prediction from experimental mass spectra.

show abstract

“…36,37 Estas novas tecnologias são úteis nas investigações de proteômicas, glicômicas, lipidômicas e genômica (ciências "omicas"), pois devido à falta de informações sobre as fragmentações, que são necessárias para a identificação de analitos, a espectrometria de massas in sílico surge como instrumento apoiador nos avanços futuros do campo de espectrometria de massas. 34,38 Assim, a construção de bancos de dados in sílico EM/EM pode complementar os resultados experimentais e ser primordial na expansão da classe de compostos analisados por esta técnica. 34 Mesmo com esforços para utilizar a química quântica na previsão e a atribuição de espectros de massas, a natureza dos processos nos quais se baseiam dificultam a reprodução sistemática e popularização desses cálculos.…”

Section: Estudos Computacionais E a Espectrometria De Massasunclassified

Estudo computacional da reatividade e do mecanismo de fragmentação de butirolactonas e derivados

Barbieri¹

View full text Add to dashboard Cite

Quero agradecer à agência de fomento CNPq, pois seu auxilio foi muito importante para o desenvolvimento de minha pesquisa; à educação pública, que foi essencial para minha formação e vem resistindo de forma corajosa ao sucateamento nos últimos anos; e ao programa de permanência estudantil, que assegurou a minha manutenção na área científica, assim como a conclusão deste projeto. Agradeço as demais agências de fomento CAPES e FAPESP pelos demais auxílios indiretos a esta pesquisa e por suas contribuições à ciência brasileira.Um agradecimento especial, também, a todos os cientistas que participaram da elaboração das vacinas contra a covid-19 e os trabalhadores da área de saúde, sobretudo aos que estavam na linha de frente de combate ao vírus. Gostaria de agradecer ao meu pai Lucrécio, à minha mãe Claudete, aos meus irmãos Diego e Marcela, que me auxiliaram e me incentivaram em todos os passos da minha vida. À minha prima Anelize, Beto e à sua filha Alice, por trazerem magia ao meu mundo. E à vó Bel, visto que sua companhia me ajudou muito, além de tornar meu dia mais agradável.Agradeço às "meninas" (Julia, Mariana, Marina, Paola e Stephanie), que me ajudaram muito, mesmo distantes, dando todo apoio emocional que eu precisava.Aos meus amigos de graduação e de pós graduação, que levarei para o resto da minha vida:

show abstract

Quantum Chemistry Calculations for Metabolomics

Cited by 60 publications

References 535 publications

An anchored experimental design and meta-analysis approach to address batch effects in large-scale metabolomics

An anchored experimental design and meta-analysis approach to address batch effects in large-scale metabolomics

MassGenie: A Transformer-Based Deep Learning Method for Identifying Small Molecules from Their Mass Spectra

Estudo computacional da reatividade e do mecanismo de fragmentação de butirolactonas e derivados

Contact Info

Product

Resources

About