Naive Bayes classification model for isotopologue detection in LC-HRMS data

Herwerden, Denice van; O’Brien, Jake; Choi, Phil M.; Thomas, Kevin V.; Schoenmakers, Peter J.; Samanipour, Saer

doi:10.1016/j.chemolab.2022.104515

“…For the latter, adducts are identi ed based on a database of frequently detected single charged adducts in LC-HRMS experiments (e.g., M+Na) [44]. Isotopes are detected based on the elemental mass defect between the parent and potential isotope mass, assuming that elemental mass defect is similar for these ions since they have the same molecular structure [45]. Whereas (in-source) fragments are further ltered based on the probability of the neutral loss (i.e., mass difference between fragment and parent ion) database, which is further elaborated in section Neutral loss database.…”

Section: Componentisationmentioning

confidence: 99%

InSpectra – A platform for identifying emerging chemical threats

Feraud

¹

,

O’Brien

²

,

Samanipour

³

et al. 2023

Journal of Hazardous Materials

Self Cite

View full text Add to dashboard Cite

“…In the past decade, a lot of efforts have been put into the generation of digital open-source/access data processing tools to tackle the complex data generated from the NTA assays [ 11 , 12 , 21 – 28 ]. These digital tools provide the means to perform a complete NTA workflow from feature detection [ 25 , 29 – 31 ] to componentization [ 21 , 31 , 32 ] and identification/annotation [ 28 , 33 – 36 ].…”

Section: Introductionmentioning

confidence: 99%

“…In the past decade, a lot of efforts have been put into the generation of digital open-source/access data processing tools to tackle the complex data generated from the NTA assays [ 11 , 12 , 21 – 28 ]. These digital tools provide the means to perform a complete NTA workflow from feature detection [ 25 , 29 – 31 ] to componentization [ 21 , 31 , 32 ] and identification/annotation [ 28 , 33 – 36 ]. These tools, even though powerful, have shown to be highly sensitive toward the data quality and the parameters used during the processing [ 9 , 37 – 40 ], particularly when dealing with complex samples [ 24 , 41 – 43 ].…”

Section: Introductionmentioning

confidence: 99%

Predicting RP-LC retention indices of structurally unknown chemicals from mass spectrometry data

Boelrijk

¹

,

Herwerden

²

,

Ensing

³

et al. 2023

Self Cite

View full text Add to dashboard Cite

Non-target analysis combined with liquid chromatography high resolution mass spectrometry is considered one of the most comprehensive strategies for the detection and identification of known and unknown chemicals in complex samples. However, many compounds remain unidentified due to data complexity and limited number structures in chemical databases. In this work, we have developed and validated a novel machine learning algorithm to predict the retention index (r$$_i$$ i ) values for structurally (un)known chemicals based on their measured fragmentation pattern. The developed model, for the first time, enabled the predication of r$$_i$$ i values without the need for the exact structure of the chemicals, with an $$R^2$$ R 2 of 0.91 and 0.77 and root mean squared error (RMSE) of 47 and 67 r$$_i$$ i units for the NORMAN ($$n=3131$$ n = 3131 ) and amide ($$n=604$$ n = 604 ) test sets, respectively. This fragment based model showed comparable accuracy in r$$_i$$ i prediction compared to conventional descriptor-based models that rely on known chemical structure, which obtained an $$R^2$$ R 2 of 0.85 with an RMSE of 67.

show abstract

“…8,13,15,[18][19][20][21] In the past decade, a lot of efforts were put into the generation of digital open-source/access data processing tools to tackle the complex data generated from the NTA assays. 12,13,[22][23][24][25][26][27][28][29] These digital tools provide the means to perform a complete NTA workflow from feature detection 26,[30][31][32] to componentization 22,32,33 and identification/annotation. 29,[34][35][36][37] These tools, even though powerful, have shown to be highly sensitive toward the data quality and the parameters used during the processing, 9,[38][39][40][41] particularly when dealing with complex samples.…”

Section: Introductionmentioning

confidence: 99%

Predicting RP-LC retention indices of structurally unknown chemicals from mass spectrometry data

Boelrijk

¹

,

Samanipour

²

,

Herwerden

³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

Non-target analysis combined with high resolution mass spectrometry is considered one of the most comprehensive strategies for the detection and identification of known and unknown chemicals in complex samples. However, many compounds remain unidentified due to data complexity and limited structures in chemical databases. In this work, we have developed and validated a novel machine learning algorithm to predict the retention index (r$_i$) values for structurally (un)known chemicals based on their measured fragmentation pattern. The developed model, for the first time, enabled the predication of r$_i$ values without the need for the exact structure of the chemicals, with an $R^2$ of 0.91 and 0.77 and root mean squared error (RMSE) of 47 and 67 r$_i$ units for the Norman and amide test set, respectively. This fragment based model showed comparable accuracy in r$_i$ prediction compared to conventional descriptor-based models that rely on known chemical structure, which obtained a $R^2$ of 0.85 with and RMSE of 67.

show abstract

Naive Bayes classification model for isotopologue detection in LC-HRMS data

Cited by 26 publications

References 16 publications

InSpectra – A platform for identifying emerging chemical threats

InSpectra – A platform for identifying emerging chemical threats

Predicting RP-LC retention indices of structurally unknown chemicals from mass spectrometry data

Predicting RP-LC retention indices of structurally unknown chemicals from mass spectrometry data

Contact Info

Product

Resources

About