Researchers from Purdue University, USA, have combined machine learning and tandem mass spectrometry to improve the flow of information in drug development.
Tandem mass spectrometry is a powerful tool used to characterise compounds in drug development. As published in Chemical Science, the team outlined two major problems in the use of machine learning in chemical sciences. The methods that are currently used do not provide a chemical understanding of the decisions that are made by the algorithm, and new methods are not typically used in blind experimental tests to see if the proposed models are accurate for use in a chemical laboratory.
Gaurav Chopra, an assistant professor of analytical and physical chemistry in Purdue’s College of Science, said: “Mass spectrometry plays an integral role in drug discovery and development. The specific implementation of bootstrapped machine learning with a small amount of positive and negative training data presented here will pave the way for becoming mainstream in day-to-day activities of automating characterisation of compounds by chemists.
“We have addressed both of these items for a methodology that is isomer selective and extremely useful in chemical sciences to characterise complex mixtures, identify chemical reactions and drug metabolites, and in fields such as proteomics and metabolomics.”
New machine learning models
The Purdue researchers created statistically robust machine learning models to work with less training data. The model looks at a common neutral reagent, called 2-methoxypropene (MOP), and predicts how compounds will interact with MOP in a tandem mass spectrometer in order to obtain structural information for the compounds.
Hilkka Kenttämaa, the Frank Brown Distinguished Professor of Analytical Chemistry and Organic Chemistry, said: “This is the first time that machine learning has been coupled with diagnostic gas-phase ion-molecule reactions, and it is a very powerful combination, leading the way to completely automated mass spectrometric identification of organic compounds. We are now introducing many new reagents into this method.”
The Purdue team introduces chemical reactivity flowcharts to facilitate chemical interpretation of the decisions made by the machine learning method that will be useful to understand and interpret the mass spectra for structural information.
