Method for Identification of Threonine Isoforms in Peptides by Ultraviolet Photofragmentation of Cold Ions

Advances, obstacles, and opportunities for machine learning in proteomics

The fields of proteomics and machine learning are both large disciplines, each producing well over 5,000 publications per year. However, studies combining both fields are still relatively rare, with only about 2% of recent proteomics papers including machine learning. This review, which focuses on the intersection of the fields, is intended to inspire proteomics researchers to develop skills and knowledge in the application of machine learning. A brief tutorial introduction to machine learning is provided, and research advances that rely on both fields, particularly as they relate to proteomics tools development and biomarker discovery, are highlighted. Key knowledge gaps and opportunities for scientific advancement are also enumerated.

Identification and Quantification of Any Isoforms of Carbohydrates by 2D UV-MS Fingerprinting of Cold Ions

Biological functionality of isomeric carbohydrates may differ drastically, making their identifications indispensable in many applications of life science. Because of the large number of isoforms, structural assignment of saccharides is challenging and often requires a use of different orthogonal analytical techniques. We demonstrate that isomeric carbohydrates of any isoforms can be distinguished and quantified using solely the library-based method of 2D ultraviolet fragmentation spectroscopy-mass spectrometry (2D UV-MS) of cold ions. The two-dimensional "fingerprint" identities of UV transparent saccharides were revealed by photofragmentation of their noncovalent complexes with aromatic molecules. We assess the accuracy of the method by comparing the known relative concentrations of isomeric carbohydrates mixed in solution with the concentrations that were mathematically determined from the measured in the gas-phase fingerprints of the complexes. For the tested sets with up to five isomers of di- to heptasaccharides, the root-mean-square deviation of 3-5% was typically achieved. This indicates the expected level of accuracy in analysis of unknown mixtures for isomeric carbohydrates of similar complexity.

Mass-Spectrometry-Based Identification of Synthetic Drug Isomers Using Infrared Ion Spectroscopy

Infrared ion spectroscopy (IRIS), a mass-spectrometry-based technique exploiting resonant infrared multiple photon dissociation (IRMPD), has been applied for the identification of novel psychoactive substances (NPS). Identification of the precise isomeric forms of NPS is of significant forensic relevance since legal controls are dependent on even minor molecular differences such as a single ring-substituent position. Using three isomers of fluoroamphetamine and two ring-isomers of both MDA and MDMA, we demonstrate the ability of IRIS to distinguish closely related NPS. Computationally predicted infrared (IR) spectra are shown to correspond with experimental spectra and could explain the molecular origins of their distinctive IR absorption bands. IRIS was then used to investigate a confiscated street sample containing two unknown substances. One substance could easily be identified by comparison to the IR spectra of reference standards. For the other substance, however, this approach proved inconclusive due to incomplete mass spectral databases as well as a lack of available reference compounds, two common analytical limitations resulting from the rapid development of NPS. Most excitingly, the second unknown substance could nevertheless be identified by using computationally predicted IR spectra of several potential candidate structures instead of their experimental reference spectra.

Spectroscopic Evidence for Peptide-Bond-Selective Ultraviolet Photodissociation

We study the photodissociation induced by ultraviolet excitation of amide bonds in gas-phase protonated peptides. Jointly, mass spectrometry and cold ion spectroscopy provide evidence for a selective nonstatistical dissociation of specific peptide bonds in the spectral region of the formally forbidden n → π* transition of amide groups. Structural analysis reveals that the activation of this transition, peaked at 226 nm, originates from the nonplanar geometry of the bond. In contrast, the statistical dissociation in the electronic ground state appears to be the main outcome of the π → π* excitation of the peptide bonds at 193 nm. We propose a tentative model that explains the difference in the fragmentation mechanisms by the difference in localization of the electronic transitions and the higher amount of vibrational energy released in the electronic excited state upon absorption at 193 nm.