HRMS platforms which have exceled at generating large amounts of high quality data have generally outpaced the delivery of fully integrated solutions which can actually mine these information-rich data sets to their maximum potential. In this talk, we will describe the latest efforts to integrate DMPK HRMS informatics workflows from acquisition to reporting. As part of a holistic strategy, we also discuss the value of adding routine use of travelling wave ion mobility and collisional cross section (CCS) data with a novel travelling wave IMS QTof informatics platform. To facilitate the routine use of large comprehensive datasets, data reduction strategies to both improve the quality of data interpretation and moderate the data burden for processing and storage will be discussed.
A variety of commercially available drug candidate molecules were incubated in microsomes or S9 fractions under typical phase I and phase II forming conditions. Compounds were chosen primarily for their ability to generate large numbers of isobaric species. Verapamil (desmethyl), warfarin (oxidation), buspirone (hydroxylation, dihydroxylation, hydroxylation with glucuronidation) were chosen as representative molecules to demonstrate the value of integrating ion mobility to a DMPK workflow. Data were compared to currently commercially available travelling wave platforms to highlight differences and similarities to the novel IMS geometry. Data for this study was collected using the most comprehensive and data rich scanning technique available, HDMSE which collects data simultaneously for low and high energy data with full ion mobility scanning information.
Through the use of a single integrated platform, we show that data can be collected routinely, processed and reviewed simultaneously across time courses and treatments. This steps away from a conventional sample-by-sample approach and allows for streamlined reviewing of much larger datasets, with navigation to the Cmax of any given metabolite in a large study with one click. In addition to accurate mass measurement for every peak, CCS values were reported giving a facile three point identification (RT, m/z, CCS). In a large percentage of cases, a two point identification (m/z, CCS) is enough to confidently discriminate isobaric species reducing the need to differentiate species purely by retention time which can be variable across large runs or across challenging matrix conditions. In addition to comprehensive three point identifications, fully integrating tools such as cross sample trendplotting, mass defect filtering, sample comparison, multivariate analysis and ion mobility data toolkits are available to interrogate all identified metabolite species and all components in the dataset. Given the magnitude of these larger datasets, a second stage of experimental work involved the implementation and evaluation of a variety of data reduction approaches. To maximize the S/N and confident detection of low abundance peaks, several novel peak detection approaches were tested which retained peaks by assessing chromatographic profiles versus random noise. In this study, this methodology could reduce the data burden by up to 20-50 fold. Depending on the purpose of the data (qualitative screening versus absolute highest sensitivity), the data file size can be tuned to meet workflow needs. A 5-10 fold data size reduction could be easily achieved with no loss in real detected peaks while also improving XIC baseline quality, spectral quality and significantly improving processing performance. This tool could be further combined with other data compression approaches.
Next generation DMPK workflows achieved through tight integration of ion mobility, HRMS, informatics and data reduction strategies.