NEW YORK (GenomeWeb) – In a paper published this week in the Journal of Proteome Research, University of Geneva researchers explored the benefits of using variable isolation windows in Swath mass spec experiments.
In an analysis of human monocyte derived dendritic cells (MDDCs), they found that use of variable isolation windows provided significant improvements in the number and quality of peptide identifications as well as improved quantitative reproducibility.
The researchers also put forth as a proof of concept a method for what they termed scheduled Swath acquisition, which, similar to scheduled selected-reaction monitoring mass spec, takes into account differences in precursor distribution across the LC elution time, allowing the instrument to optimize the timing of mass spec cycles to analytes' LC retention times.
In Swath-style data independent acquisition methods, mass spectrometers cycle through broad m/z isolation windows, fragmenting all precursors in those windows, which allows the instrument to collect MS/MS spectra on all the ions in a sample.
The first Swath algorithms, including the original commercial algorithm released by AB Sciex in 2011, used set isolation windows of 25 m/z throughout the entire mass spec cycle. However, a sample's ions are not evenly distributed throughout the m/z range. Rather, there are typically fewer ions at the lower and higher m/z ranges and more ions in the middle band of m/z ranges.
The more ions in a given isolation window, the more there will be overlapping and interfering spectra, making for more difficult data analysis and reduced sensitivity. One way to address this issue is to use smaller windows for busier portions of the m/z range, reducing the complexity of the resulting data.
This notion has led several vendors in recent years to release Swath-style algorithms that, instead of using set isolation windows throughout the entire cycle, allow researchers to vary the window size as they move through the m/z range. For instance, both AB Sciex's Swath 2.0 and Thermo Fisher Scientific's pSMART-DIA allow for use of variable windows.
In their JPR paper, the Geneva team, which was involved in development of both the original Swath and Swath 2.0 algorithms, put forth their own tool for implementing variable isolation windows in Swath-style experiments.
Named SwathTuner, the tool is meant as an open-source, vendor-independent option for doing Swath experiments with variable isolation windows, University of Geneva researcher Emmanuel Varesio, senior author on the JPR paper, told GenomeWeb. Additionally, it offers what he said was a more "user-friendly interface" and more extensive visualization tools than AB Sciex's Swath 2.0 tool.
Applying the method to analysis of MDDCs, Varesio and his colleagues were able to identify on the order of 10 percent more peptides and 10 percent more proteins using variable windows than with a fixed window approach. Specifically, when they based their isolation window sizes on the precursor ion population (PIP) of a window, they identified 13.8 percent more peptides and 13.1 percent more proteins than they did with a fixed window approach. When they based their window sizes on the total ion current (TIC) within each window, they identified 8.4 percent more peptides and 10 percent more proteins. Using the Swath 2.0 tool, they identified 9.5 percent more peptides and 9.5 percent more proteins.
They also found that variable windows improved their quantitative reproducibility, with 92.2 percent, 92.1 percent, and 93.3 percent of total peptides having CVs of under 20 percent when using, respectively, the PIP, TIC, and Swath 2.0 approaches. Using fixed windows, 84.9 percent of peptides had CVs under 20 percent.
The Geneva team also put forth in the paper the notion of scheduled Swath, which allowed them to use different sets of variable windows for each LC retention time segment to account for differences in precursor distribution throughout the full LC run.
"During the LC gradient, the ion population is shifting," Varesio said, noting that this means that the optimal combination of window sizes will vary throughout the LC gradient with, for instance, windows that were optimally narrow at the beginning of the gradient becoming broader toward the end of the gradient.
The researchers do not have a finished algorithm for such scheduled Swath, but, in a proof of concept exploring the idea they found that it improved data quality. Varesio said, however, he was uncertain whether they would follow up with development of a formal algorithm for the approach.
To an extent, he said, it would depend on vendor interest, as implementing it would involve hardware modifications.