Vendor Centroiding Is Not Supported For Waters Ion Mobility Data

Why Your Waters Ion Mobility Data Won't Centroid — And What to Do About It

You spent the last six hours running a batch of ion mobility samples on your Waters instrument. Still, if that's happened to you, you're not alone, and you're not doing anything wrong. Plus, the raw files are sitting on your drive, full of promise. Worth adding: you fire up your favorite DIA analysis pipeline, point it at the data, and — nothing. Or worse, you get an error message telling you that vendor centroiding is not supported for Waters ion mobility data. This is a real, documented limitation that catches a lot of people off guard, especially when they're bridging the gap between vendor-specific workflows and open-source or third-party analysis tools.

Let's break down exactly what's going on, why it happens, and — most importantly — what you can actually do about it.

What Is Vendor Centroiding and Why Does It Matter?

Understanding Centroid Data vs. Profile Data

Mass spectrometers produce data in two fundamental forms: profile mode and centroid mode. Here's the thing — profile data is continuous — it records the full shape of every ion signal across the mass range, like a smooth landscape of intensity values. So naturally, centroid data, on the other hand, is a simplified representation. Which means the software has already identified the peaks and recorded only their center m/z values, intensities, and widths. Think of it as the difference between a photograph and a set of coordinates marking where the bright spots are.

Centroiding is the process of converting from profile to centroid. It's a preprocessing step that many analysis workflows require because centroid data is smaller, faster to process, and easier to work with for certain types of quantification and identification algorithms.

What Makes Waters Ion Mobility Data Different

Waters instruments that perform ion mobility spectrometry — such as the SYNAPT family, the SELECT SERIES Cyclic IMS, and related platforms — add a mobility dimension to the traditional m/z and intensity data. So the result is a three-dimensional dataset: m/z, intensity, and drift time (or collision cross-section). This extra dimension is incredibly valuable for separating isobaric species and improving peptide identification rates, but it also complicates data handling.

The raw data format from Waters instruments — typically .So raw files using the proprietary Waters format — encodes this ion mobility information in a way that's tightly coupled to the vendor's own software ecosystem, like MassLynx or UNIFI. The internal structure of these files, particularly how the ion mobility spectra are stored and indexed, is not something that open-source or third-party tools have been able to fully replicate or parse for centroiding purposes.

What "Vendor Centroiding Is Not Supported" Actually Means

When you see that message, it's telling you that the analysis software you're using doesn't have a built-in capability to read the Waters raw files, extract the ion mobility dimension, and convert the profile data into centroid format on your behalf. The software either doesn't support Waters ion mobility files at all, or it supports them only in profile mode — which, depending on your workflow, might not be usable.

We're talking about distinct from standard Waters mass spectrometry data without ion mobility, where centroiding support may or may not be available depending on the tool. The ion mobility layer adds a complication that most general-purpose tools haven't addressed yet.

Why This Limitation Exists

The Complexity of the Waters Raw Format

Waters doesn't publish a complete specification of their raw file format. Because of that, the internal structure is proprietary, and it changes between instrument generations and software versions. Reverse-engineering it is a massive undertaking, and the ion mobility component makes it even harder. The mobility spectra are stored alongside the conventional mass spectra, but the relationship between them — how they're aligned, how the drift time calibration works, how frames are organized — isn't documented in a way that makes it easy for external developers to build reliable parsers.

The Niche Audience Problem

Ion mobility-mass spectrometry is a powerful technique, but it's still a relatively specialized application compared to standard LC-MS or GC-MS. The number of people who need to process Waters ion mobility data outside of the vendor's ecosystem is smaller than the number who need to process, say, Bruker or Thermo raw files. That smaller audience means fewer resources have been devoted to building strong, maintained parsers for this specific combination.

Different Priorities in the Open-Source World

Many open-source DIA tools were built with Thermo, Bruker, or SCIEX data in mind first. Waters support — even for non-ion-mobility data — has historically been a secondary concern. When you add ion mobility to the mix, the development priority drops even further. The maintainers of these tools are often volunteers or small teams with limited bandwidth, and they tend to prioritize the formats that serve the largest user base.

How This Affects Your Workflow

The Pipeline Breaks at the First Step

If your analysis workflow depends on centroided data — and many do, particularly for targeted DIA methods using tools like OpenSWATH or certain Skyline configurations — you hit a wall before you even get to the interesting part. The data conversion step fails, and nothing downstream can run.

You're Forced Into the Vendor's Ecosystem

The most straightforward workaround is to stay inside MassLynx or UNIFI for everything. You can export results, but the flexibility to use your preferred analysis tools is gone. For labs that have invested in open-source pipelines or that need to integrate Waters ion mobility data with data from other instrument platforms, this is a real constraint.

Data Sharing Becomes a Bottleneck

If you're collaborating with someone who uses a different platform, or if you're depositing data in a repository, the inability to convert Waters ion mobility files into a more universal format adds friction. Standard formats like mzML can store ion mobility data, but the conversion from Waters raw files to mzML with proper ion mobility preservation isn't always straightforward when centroiding is involved.

Workarounds and Practical Solutions

Use the Vendor's Export Tools First

One pragmatic approach is to use MassLynx or UNIFI to export your data into a format that your downstream tools can handle. This might mean exporting to mzML, mzXML, or even a centroided text-based format. Also, the trade-off is that you're doing the conversion inside the vendor software, which means you're trusting their algorithms for peak picking and centroiding. That's usually fine — MassLynx's centroiding is mature and well-tested — but it does mean you lose the ability to tweak centroiding parameters in your preferred tool.

If you found this helpful, you might also enjoy the 3 particles of an atom are or the electrons in the outermost energy level of an atom.

Try the MSConvert Route

ProteoWizard's MSConvert is the Swiss Army knife of mass spectrometry file conversion, and it supports a wide range of input formats. In some cases, MSConvert can read Waters raw files and convert them to mzML with ion mobility data preserved. That said, whether centroiding is applied during this conversion depends

When you invoke MSConvert, the default behaviour is to apply the vendor’s own centroiding algorithm to the incoming profile data. If you need a true centroided representation — especially for downstream tools that expect a single m/z value per scan — you must explicitly request it. In real terms, the -c flag forces MSConvert to collapse each scan into its most intense pixel, preserving the ion‑mobility dimension as an additional metadata field. Conversely, omitting this flag leaves the original full‑waveform data intact, which can be advantageous when you intend to re‑centroid later with a method that lets you control parameters such as signal‑to‑noise threshold or peak width.

Because Waters raw files embed ion mobility as a series of “drift‑time” values rather than as a separate mobility scan, the conversion process must decide how to treat those values. In practice, two strategies dominate:

1. Preserve the raw drift‑time array and write it into the mzML container as a user‑defined mzML attribute (e.g., MS:1001083). This approach keeps the original resolution of the mobility dimension, allowing later analysts to bin, smooth, or deconvolute the data according to their own protocols. Even so, many open‑source tools still expect ion mobility to appear as a distinct scan number; in those cases the attribute may be ignored, effectively rendering the mobility information invisible.

2. Apply on‑the‑fly centroiding during conversion. By enabling the -c option together with the -mzML output flag, MSConvert will generate a list of centroids (one per scan) while still embedding the drift‑time array as a separate field. The resulting mzML file can then be opened in software such as Skyline, OpenMS, or MS‑Explorer, which read the centroid values and the associated mobility metadata without further manipulation.

If you find that MSConvert consistently drops the ion mobility column, a more manual route is to use the Waters WiffToolkit (or its open‑source counterpart wiffreader) to extract the raw profile data, then feed those profiles into a dedicated centroiding script written in Python or R. Libraries such as pyms or mzR let you define a custom centroiding routine that respects the ion‑mobility timestamps, ensuring that the final mzML reflects exactly the same arrival‑time distribution as the original instrument output.

Another avenue worth exploring is MS‑Convert’s “-s” (scan‑selection) mode. By specifying a subset of scans — e.g.Practically speaking, , only those that contain the ion mobility data — you can avoid converting irrelevant profile scans that would otherwise bloat the mzML file and complicate downstream parsing. This is particularly useful when a single Waters RAW file contains a mixture of full‑scan MS¹, MS², and ion mobility scans.

Beyond conversion, the community has begun to standardise the representation of ion mobility in mzML. Practically speaking, the recent addition of the MS:1001083 “Ion Mobility m/z” type and the MS:1001084 “Ion Mobility drift time” fields, defined by the PSI‑MFG working group, provides a clear schema for attaching drift‑time values to each centroid. When MSConvert (or any other converter) respects these tags, downstream applications can automatically recognise and utilise the mobility dimension without additional mapping steps.

Practical recommendations for a smooth workflow:

• Export from the vendor first if you need a quick, reliable centroided file; MassLynx’s built‑in export to mzML is mature and typically preserves ion mobility metadata when the “include mobility” option is enabled.
• Validate the conversion by loading the resulting mzML into two independent viewers (e.g., Skyline and OpenMS) and confirming that the drift‑time values match the original RAW file’s header information.
• Automate the process with a small wrapper script that calls MSConvert with the appropriate flags, then checks the presence of the ion mobility fields using the mzML Python library. This reduces manual error and makes the pipeline reproducible.
• Consider alternative converters such as mzML2File or MSiReader, which have been updated to handle Waters RAW files with ion mobility support, especially when the -c flag is combined with explicit drift‑time extraction.

By aligning the conversion settings with the expectations of your downstream analysis tools, you can avoid the common “pipeline break” that occurs at the very first step. The key is to treat ion mobility not as an afterthought but as a first‑class citizen of the data format, ensuring that both centroiding decisions and metadata preservation are made consciously rather than left to default behaviours.

Conclusion

Ion mobility adds a valuable dimension to modern mass‑spectrometric workflows, yet its integration often stalls because the necessary conversion steps are neither obvious nor uniformly supported across tools. Worth adding: as open‑source converters continue to adopt the PSI‑MFG ion mobility tags and as more laboratories demand seamless data exchange, the ecosystem is moving toward greater interoperability. The challenges — ranging from loss of centroided information to incomplete preservation of mobility metadata — can be mitigated through deliberate use of vendor export options, strategic MSConvert flags, and, when needed, custom scripting that respects the unique structure of Waters RAW files. Until those refinements become universal, adopting a disciplined conversion strategy — validated across multiple software platforms — will keep your ion mobility data usable from the first step onward, enabling the full analytical potential of these high‑resolution measurements.