Let the Data Flow
Rethink your data analysis tools for flow cytometry.
Since its invention in the 1960s, flow cytometry—a technique used to identify and sort specific populations of cells—has extended its reach beyond immunol-ogists to those performing diverse assays in both the basic and clinical research arenas. Whether you are using the technology to detect rare stem cells or to diagnose blood disorders, the vast improvements in the machines themselves in recent years has upped the need for robust software. “With the new instrumentation that’s coming out, it is just going to encourage people to collect more data,” says Ryan Duggan, technical director of University of Chicago’s flow cytometry core facility. For example, instruments can now detect 20,000 “events”—such as a cell or piece of debris—in a single second, and can detect cells labeled with nearly 20 different fluorochromes in a single run.
The majority of flow cytometry users settle into one type of software that they’re most comfortable with. As a result, “they don’t necessarily explore everything the software can do,” says Derek Davies, who runs a fluorescence-activated cell sorting laboratory at the London Research Institute, Cancer Research UK.
Whether you’re using software tools provided by a core facility or looking to buy your own, that extra functionality may ease your analysis. And there are a lot of options to choose from. Because flow cytometry data files aren’t locked up in a proprietary file format, many third-party companies have entered the market to address common problems flow users encounter, such as visualizing complex multidimensional data, having to repeat the same analysis on different batches of data, and struggling with slow lag times in manipulating large datasets.
To help you identify which program is best for your data, The Scientist profiles a few data analysis software tools on the market.
Overview: For offline data analysis—that is, analysis after rather than during data collection—FlowJo dominates the market and is a favorite among core facility managers and researchers alike. The program contains a “workspace” window with a view of all samples and their analyses, in which users can organize and manage their experiments and produce reports.
Best for: Core facilities or individual users who want all the bells and whistles
System requirements: For Mac version 9.0, at least 512 MB of RAM is needed, and OSX 10.4 is highly recommended. For PC version 7.6, you’ll need Java 1.6.
• FlowJo offers a broad range of analyses, including cell-cycle analysis, proliferation, and kinetics. Overall, “it’s the best software we can find in the current time,” says Yanhui Deng, director of cell sorting services at Boston University Medical Center’s flow cytometry core facility.
• It’s one of the few Mac-compatible offline analysis tools.
• It has good automated and manual post-acquisition color compensation, which accounts for the spectral overlap in cells labeled with multiple types of fluorochromes.
• FlowJo is not intuitive for beginners, Duggan says, though the company offers training sessions and support.
• It is slow with large or multiple datasets. For example, to sort stem cells, which might be present in less than 1% of your population, you’ll end up running at least 5 to 6 million cells through the fluidic system and past the lasers. The program lags with more complex analyses on that amount of datapoints.
• It offers much more than what the average flow cytometry user needs, and there is no “lite” version.
• The 3D data viewer works well but is currently only available in the Windows version.
Cost: $2,495 for commercial purchase, and $1,495 for academic labs for a single license. Site license price varies based on the number of computers included. Upgrades from previous versions range from $499 to $799.
Overview: De Novo Software released the first version of FCS Express in 1999. Now on version 3.0, this offline software is known for being approachable and highly customizable.
Best for: Beginner to advanced flow cytometry users
System requirements: For version 3.0, you’ll need Windows NT/2000/2003/XP/Vista/7 and at least 256 MB of RAM.
• It’s user friendly for anyone who has worked with Microsoft Office, and researchers can easily generate PowerPoint slides of their data.
• FCS 3 Express Reader, which is free, allows you to share your gating strategy (instructions to the program about which cell populations to include or exclude) and other analyses with collaborators.
• A “lite” version lets casual users analyze experiments with one or two fluorochromes instead of 18.
• A display slide within the program (separate from the analysis window) helps prepare data for a paper or talk. You can link statistics in this slide to your analysis files, so if you go back and alter your analysis, the stats on your display slides will automatically change.
Cons: FCS Express is not as feature rich as FlowJo. It does have cell cycle and proliferation add-ins, but does not have as many high-level applications. “As a result the FCS is more streamlined, and is for researchers who need to go from robust data analysis to a PowerPoint slide,” Duggan says.
Cost: For the Version 3 Lite Research Edition, $300 to $500. For the Professional Research Edition, prices vary based on the number of licenses purchased. A single copy containing two licenses costs $1,600 for academic labs.
Overview: Kaluza (Beckman Coulter) was introduced late last year and is in the hands of a relatively small group within the flow cytometry community. This offline software offers new ways to analyze and visualize large, complex datasets.
Best for: A slightly younger generation used to seeing Web-based apps or FlowJo users who have a lot of data, Duggan says.
System requirements: For PC, Windows XP and Vista 32-bit. For Mac, operating systems under dual-boot programs such as “Bootcamp.”
• It’s fast and adept at handling large data files. When you redraw a gate, the software responds instantaneously, Duggan says.
• User friendly and aesthetically pleasing, with a minimalistic look. It mimics the Microsoft Office Suite with a ribbon of different functions at the top of the window.
• Tree-plots allow users a different way to see multi-dimensional data. (This feature is also present in WinList, below.)
• Doesn’t have the cell cycle or kinetics functions built in yet. “At this point, it’s similar in capabilities to FCS Express without the add-ins, or standalone WinList,” says Duggan. That said, it will fulfill most users’ needs: gating, generating stats, and processing batches of data.
• Words used within the program—such as “panels,” “protocols,” and “tests”—are geared for clinicians and don’t always translate well to basic research, according to Duggan.
Cost: Single user licenses range from $2,000 to $3,000, depending on the specific kit you purchase. Multi-user network versions go for around $2,000 or less per license.
Overview: One of the original flow cytometry software tools, WinList (Verity Software House) has been around since 1991. It’s known for its computational strength. One offshoot of WinList, a relatively new program called Gemstone, tackles automated rather than user-guided analyses.
Best for: Flow cytometry experts, Duggan says.
System requirements: Windows 2000 or higher
• You can easily generate a macro containing a sequence of analysis steps, thus speeding up repeat analyses.
• The tool comes with powerful statistics, such as Chi-squared population comparisons.
• The program is good at taking complex datasets and simplifying them into a single plot that has the most information packed into it.
• It has a steeper learning curve than more basic software tools, like FCS Express.
• WinList does not come with cell cycle analysis, but Verity Software House also makes ModFit, the go-to software for this application.
Cost: The prices vary depending on licensing, but a primary user license is $1,850. It’s $400 for an upgrade from a previous version.
Overview: FACSDiva is a data acquisition and analysis software. The latest version, 6.0, has been available since last year and comes with most BD flow cytometers.
Best for: Individual labs that purchase a BD instrument for basic flow cytometry analysis
System requirements: For version 6.0, 2 GB of RAM if connected to the flow cytometer. If installing on Windows XP, make sure your operating system has Service Pack 2.
• It’s a timesaver to learn only one software package, instead of two separate packages for acquisition and analysis. And FACSDiva is one of the best acquisition tools available, Duggan says.
• After the acquisition step, it’s easy using this program to transform data using various algorithms to better see and define cell populations.
• The software is well supported, with bug fixes that are quickly addressed.
• It lacks some basic functions, such as allowing users to overlay plots from various runs.
• The software does not offer users more advanced tools, like cell cycle analysis, but data can easily be exported for offline analysis.
Cost: A single user license is $3,195.
Don’t forget freeware. WinMDI is another popular choice, although it is no longer supported. “It does very basic analysis, and that’s maybe all you want initially,” Davies says. WinMDI is not compatible with some of the newer flow cytometry instruments. Other freebies include FlowingSoftware, WEASEL and open-source flow cytometry tools within Bioconductor.
Plan experiments better. Free, Web-based tools by Invitrogen and Becton Dickinson are making it easier for researchers to smartly choose multiple fluorescence markers that avoid significant overlap in the spectra in which their markers are detected. These tools may reduce the time users spend in the data analysis step in which they computationally adjust for spectral overlap.
Rejoice in automatic gating. Experts say data analysis software is moving toward automatic gating. FlowCAP, a project of the International Society for Advancement of Cytometry, was created last year to spur computational advancements for gating. This can save researchers hours of manually setting their boundaries, notes Ryan Brinkman, a researcher from the University of British Columbia and member of the FlowCAP committee. “In the future, people will spend their time focusing on why these cells are different rather than where the line should go between the two populations,” he adds.