Science depends on teamwork and rapid sharing of results. Tools like the Global Natural Products Social Molecular Networking (GNPS) platform have changed the way mass spectrometrists approach data analysis and collaboration. GNPS lets you process and share your data and molecular networks with lab mates, supervisors, and reviewers.
A collaborative team of researchers have introduced a new tool that streamlines the sharing, review, and understanding of contextual mass spectrometry data: the GNPS Dashboard. We interviewed Mingxun Wang from the University of California at San Diego to learn what the GNPS Dashboard is, how they developed it, and how you can use it to streamline collaboration.
COVID-19 disruptions highlighted the need for new tools
Before March 2020, sharing data with your coworkers or supervisor was easy: they just looked over your shoulder. Together, you could browse the data, find ions of interest, and investigate their distribution across samples. But as labs shut down and researchers began working remotely, this became complicated.
Many research labs weren’t ready. And neither were their IT departments. Problems with remote access like configuring VPNs and adhering to licensing agreements from software vendors quickly became bottlenecks in routine analyses. Worse, many programs that researchers used to browse their data simply wouldn’t work on their home operating system.
Like many scientists adjusting to new COVID-19 protocols, Ming understood this all too well: “I work on several different machines with different operating systems, and I’m not going to have my processing software on all of them. Even though there are cross-platform solutions, it just never was as seamless as it should be.”
“We send links back and forth for analyses – like jobs you run in GNPS. So, it feels obvious that if you wanted to look at the raw data, it should be as simple as clicking on a link and everything pops up. So that’s really where it started – we have this data; shouldn’t it be easy to share it?”
A web-based way to browse and share your data
The GNPS Dashboard removes many of the barriers to remote collaboration. It’s designed to integrate with the data formats already used in MassIVE. So you can simply retrieve raw data and begin analysis from any browser. There’s no software to install, so no problems with operating systems. In fact, these problems are so pervasive that even established software companies like MestreNova and ACD have both released cloud-based processing options recently.
A useful tool for reviewers
The GNPS Dashboard provides direct access to the data at any point in your analysis. At any time, you simply click a button to generate a unique URL for the visualization you’ve processed. Reviewers then simply click the link you provide and see exactly the same chromatograms.
For team projects, you can send links to serve as quality checkpoints or simply highlight new information. With the Dashboard, you don’t have to walk your colleagues through every step of the data processing on their machine.
Data transparency is a hot topic. Reviewers routinely request that raw data is made publicly available to allow independent verification of published results. Increasingly, researchers deposit their data on central databases like MassIVE, the Metabolomics Workbench, or in archives like Zenodo to allow reanalysis and validation.
As Ming points out, “If it’s expected to look at the data as part of the review process, it creates a reason for the authors to make their data public. But before, if they made their data public, there’s a chance that nobody actually looked at it because it's a pain.”
Now, it’s as easy as sharing a link in the supplemental information. It’s a huge relief when a collaborator or reviewer says, “We’re able to see the data and validate your claim,” Ming observes.
Mass spectrometry in a hands-on classroom
The GNPS Dashboard has a major advantage over local programs: it operates in real time between concurrent users. Using a synchronized leader/follower mode, collaborators and classrooms can walk each other through the data directly.
“We had always thought the dashboard was going to be for two people, three people tops.” Ming explains, “Then I got an email from a collaborator who was asking ‘is there a way to install software easily to look at MS data in the classroom?’ and I said ‘Do I have a tool for you!’”
Teaching students and new researchers about the pitfalls of MS analysis can be challenging. Hands-on demonstrations are important to connect textbook lessons on theory to the actual data we work with. MS data are never perfect. And there are numerous common pitfalls that an advisor can point out as they walk through data with a student or class.
“Exposing the data allows for problems to be identified – and from a long-term perspective – this increases education and awareness of how to correct it.”
As Ming puts it, an advisor can go further into the analysis with the student: “Since you brought me these data, this is the pattern in the data that is incorrect.” Mentors can walk students through concepts at all layers, like common isotope patterns, real chromatography examples, or problems in the fragmentation patterns.
In workshops and large classrooms, lecturers walk audiences through step-by-step processing while explaining what boxes and buttons to click, just to get them to the next step. In contrast, the GNPS Dashboard enables a more guided approach:
“A student can be carried through steps 1-10, and then have the ability to explore. They don’t get lost because they’re pulled along as a follower, but then they detach and can go on to do step 11.”
When it comes to scale, the GNPS Dashboard can even handle large classes. “These days, we've been scaling it to a couple hundred users at the same time in our workshops.”
Using the GNPS Dashboard
GNPS molecular networking offers direct context about a feature – through its relatedness to other features – within a set of samples. The GNPS Dashboard builds on this by allowing you to take an important feature from your network and directly view its distribution, intensity, and chromatography profile.
You can either browse for data using the GNPS Dataset Explorer or upload your data directly to the interface to begin analysis. From there, you can easily view the total ion or base peak chromatograms (TIC/BPC). Or, if you’re doing targeted analysis, you can simply provide a mass and view the extracted ion chromatogram (XIC).
The panels are interactive, so when you click on the TIC window, you see the MS1 spectrum for any time point. Using the heat map feature, you can see when MS2 data was acquired throughout the run and select a scan to view the fragmentation.
When initiating a leader/follower session, simply click the “Sync Options” button and provide a collaborator with the session ID they need to connect to your session. From there, you can explore the dataset together in real time.
Innovation through iteration
Early users of the standard GNPS molecular networking know that the platform’s improvements over the last eight years have drastically improved how they view their data.
“It's that drive – having interoperability and interconnectedness – where you can have new tools that are properly contextualized to drive users, at the moment that they might need it,” says Ming.
“One of the coolest parts is that we ended up making it more extendable than we originally planned. We integrated other GNPS tools and things like mzMine for proteomics feature finding.”
Ming’s excitement for what comes next is contagious. And he’s quick to point out that the big improvements come directly from community feedback and collaboration.
“If you have a tool that can integrate into the GNPS Dashboard that can easily visualize the results and enhances the usability of the dashboard itself, work with us and we can integrate it.”
To learn more about the GNPS Dashboard:
- Check out the recent paper in Nature Methods;
- See the highlighted use cases for how others are using it;
- Grab some data from the GNPS Dataset Explorer and start exploring!
Working on new LCMS analysis tools?
If you’re developing new tools for LCMS analysis, we’re experts at scaling and optimizing them for community use. Schedule a call to learn how we can optimize your tool.