Prediction of MS/MS Data. 1. A Focus on Pharmaceuticals Containing Carboxylic Acids

Publisher Website
Google Scholar
Abstract
Metabolite identification is a necessary step in developing safe and effective drugs. Metabolite analysis typically involves rapid identification of the chemical composition of the metabolite by automated HPLC−MS methods, followed by the laborious process of identifying the structure of the metabolite. Since MS is typically utilized to identify the metabolite, it is logical to utilize MS/MS to structurally characterize the sample. However, interpretation of MS/MS data may not provide sufficient information, as fragmentation pathways are not well understood or predictable. Therefore, other more time-consuming methods of analysis are often undertaken. If the dissociation rules for low-energy MS/MS experiments were clearly defined for all classes of compounds, more information would be obtained from MS/MS data, and metabolite identification would proceed more rapidly. We are currently developing methods to define these fragmentation rules. By screening ∼100 carboxylic acids at a time and applying knowledge of physical-organic chemistry, predictive rules are under development that describe how compounds dissociate under low-energy collision-induced dissociation conditions. Studies of carboxylic acid dissociation demonstrate that this approach is practical and reliable. Dissociation rules were predicted with a 90% success rate, when tested on acid-containing pharmaceuticals. This predictive power cannot be matched by any commercially available software. This study, and others like it, will be used to develop algorithms that more rapidly identify drug metabolites and degradation products, based on MS/MS data. Such algorithms will benefit drug development for all types of pharmaceuticals.

This publication has 20 references indexed in Scilit: