209
OMICS A Journal of Integrative Biology
Volume 10, Number 2, 2006
© Mary Ann Liebert, Inc.
Data Standards for Flow Cytometry
JOSEF SPIDLEN,1 ROBERT C. GENTLEMAN,2 PERRY D. HAALAND,3
MORGAN LANGILLE,1 NOLWENN LE MEUR,2 MICHAEL F. OCHS,4
CHARLES SCHMITT,3 CLAYTON A. SMITH,1 ADAM S. TREISTER,5
and RYAN R. BRINKMAN1
ABSTRACT
Flow cytometry (FCM) is an analytical tool widely used for cancer and HIV/AIDS research,
and treatment, stem cell manipulation and detecting microorganisms in environmental sam-
ples. Current data standards do not capture the full scope of FCM experiments and there
is a demand for software tools that can assist in the exploration and analysis of large FCM
datasets. We are implementing a standardized approach to capturing, analyzing, and dis-
seminating FCM data that will facilitate both more complex analyses and analysis of datasets
that could not previously be efficiently studied. Initial work has focused on developing a
community-based guideline for recording and reporting the details of FCM experiments.
Open source software tools that implement this standard are being created, with an em-
phasis on facilitating reproducible and extensible data analyses. As well, tools for electronic
collaboration will assist the integrated access and comprehension of experiments to empower
users to collaborate on FCM analyses. This coordinated, joint development of bioinformat-
ics standards and software tools for FCM data analysis has the potential to greatly facili-
tate both basic and clinical research—impacting a notably diverse range of medical and en-
vironmental research areas.
This paper is part of the special issue of OMICS on data standards.
INTRODUCTION
FLOW CYTOMETRY (FCM) is a technique used in basic and clinical research for studying the immuno-logical status of patients treated with vaccines or other immunotherapies, for characterizing cancer,
HIV/AIDS infection, and other diseases, as well as for research and therapy involving stem cell manipula-
tion (Braylan, 2004; Hengel et al., 2001). It is also used for studying environmental samples, such as de-
tecting specific mocroorganisms in soil or water samples (Lomas, 2004) (Fig. 1). In FCM, intact cells and
their constituent components are tagged with fluorescently conjugated monoclonal antibodies and/or stained
with fluorescent reagents and then analyzed individually. In the instrument, hydrodynamic forces line cells
1Terry Fox Laboratory, British Columbia Cancer Research Center, Vancouver, Canada.
2Fred Hutchinson Cancer Research Center, Seattle, Washington.
3BD Technologies, Research Triangle Park, North Carolina.
4Fox Chase Cancer Center, Philadelphia, Pennsylvania.
5Tree Star, Inc., San Carlos, California.
6204_19_p209-214 7/24/06 1:54 PM Page 209

up in single file and the fluorescent molecules in/on each cell are excited by laser light at speeds that can
exceed 70,000 cells per second (Bonetta, 2005). The fluorescence emission from each cell is collected by
a series of photomultiplier tubes, and the subsequent electrical events are collected and analyzed on a com-
puter that assigns a fluorescence intensity value to each signal in Flow Cytometry Standard (FCS) data files
(Seamer et al., 1997). FCM analysis involves identifying intersections or unions of polygonal regions in
hyperspace that are used to filter or “gate” data and define a subset or sub-population of events (or exclude,
for example, cell debris) for further analysis or sorting.
The International Society for Analytical Cytology (ISAC) has adopted the FCS Data File Standard for the
common representation of FCM data. This standard is supported by all of the major analytical instruments to
record the measurements from a sample run through a cytometer, and scientists can choose among instruments
and software with no major data compatibility issues. However, this standard stops short of describing the pro-
tocol used or the computational post-processing and data analysis performed in an FCM experiment. It also
does not cover data interpretation, one of the most difficult and time-consuming aspects of the entire analyti-
cal process (Braylan, 2004). FCM has traditionally been a manually intensive technique; however, automated
high-throughput FCM techniques have been recently developed that can rapidly collect large data sets with
complexities similar to gene microarrays (Gasparetto et al., 2004). The huge amount of information generated
by high-throughput technologies need to be transformed into executive summaries that are brief enough for
creative studies by a human researcher (Brazma, 2001). One of the most insidious problems in accomplishing
this goal is the lack of standard data formats for information exchange (Chicurel, 2002). One basic challenge
for FCM is to greatly simplify, from the end user’s viewpoint, data analysis and extraction of statistical infor-
mation (Boddy et al., 2001; Herzenberg et al., 2002). This needs to happen in a highly systematic, automated,
and traceable way that retains flexibility and is consistent with current visually implemented tools. Further re-
quirements include organizing data in such a way that raw data can be combined from multiple centers and
scientists and clinicians at remote locations can collaborate on data interpretation. Such needs are currently lag-
ging behind the ability to actually collect the samples and run the FCM analyses (De Rosa et al., 2003).
DATA STANDARDS METHODOLOGY AND GOALS
To address these shortcomings, we responded to the NIH Program Announcement “Innovations in Bio-
medical Computational Science and Technology” (PAR-03-10), which solicited proposals for the develop-
ment of new informatics, computational and mathematical tools and technologies including platform-inde-
pendent translational tools for data exchange and for promoting interoperability. We have brought together
a cross-disciplinary international collaborative group of bioinformaticists, computational statisticians, soft-
ware developers and clinician scientists, from both academia and industry (including both software and hard-
ware suppliers) to collaborate on the development of data standards for flow cytometry. In conjunction with
the ISAC data standards committee and an Institute of Electrical and Electronics Engineers (IEEE) Work-
ing Group (Bioinformatics Standards for Flow Cytometry WG, P1943.2), our goal is to provide consistency
in the electronic recording of flow cytometry data analysis. We aim to create universal solutions for repre-
senting, collecting, annotating, archiving, analyzing and disseminating flow cytometry data and analyses.
DATA STANDARDS FOR FLOW CYTOMETRY
211
FIG. 1. An example of flow cytometry analysis: Sargasso Sea samples at different depth. Chlorophyll fluorescence
versus forward scatter of marine microbes (mainly the cyanobacterium Prochlorococcus—a microorganism approxi-
mately 0.2 m in diameter) from samples taken at different depths at the Bermuda sampling station (BATS program;
directed by Michael Lomas of the Bermuda Biological Station for Research). Chlorophyll fluorescence corresponding
to the relative chlorophyll content was detected using a filter for 672–712 nm. Forward-scattered light (FSC) is pro-
portional to cell-surface area or size. FSC is a measurement of mostly diffracted light and is detected just off the axis
of the incident laser beam in the forward direction. Measurements were obtained using a Cytopeia Influx jet in air
sorter, 200 mWatt excitation by a 488 nm laser. The plots indicate that chlorophyll content and size of these marine
organisms remain constant, with increasing the depth up to 100 m where they both increase and up to 200 m where
the populations disappear. The small tight cluster that appears in some of the panels (indicated by arrow in 200-m sam-
ple) represents 0.5-m calibration beads.
6204_19_p209-214 7/24/06 1:54 PM Page 211

First, we are establishing a guideline that outlines the minimum information required to unambiguously
record, report, interpret, and reproduce FCM experiments, the Minimum Information for a Fluorescent Ac-
tivated Cell Experiment (MI-FACE), to promote the standardized documentation of experimental details
(Fig. 2). In developing this guideline, we are, in part, capitalizing on previous bioinformatics standards de-
velopments, most notably the development of the Minimum Information About a Microarray Experiment
(MIAME) standard (Brazma et al., 2001; Ball and Brazma, this issue), a process that is being successfully
adopted by other communities (Le Novere et al., 2005; Fiehn et al., 2006, Taylor et al., this issue). This
guideline will be encapsulated using the Unified Modeling Language (UML) and a model will be created
to enable its application within various software components using the Extensible Markup Language (XML)
standard. It is also necessary to develop platform independent extensions of the model for describing com-
pensation and gating to enable the cross-platform comparison of gating results and analytical pipelines. In
order to ensure collaboration, it is also essential to provide the standardized use of the data attributes though
documentation and controlled vocabularies (Taylor et al., this issue; Orchard et al., 2005).
CURRENT STATUS
Though we are still within the first year of this project, we have developed the first draft of a Fluores-
cent Activated Cell Experiment Ontology (FACE Ontology), including controlled vocabulary for referring
to terms. We built on the knowledge encapsulated in CytometryML (Leif et al., 2003), an XML encapsu-
lation of FCM metadata, transforming CytometryML into the World Wide Web Consortium (W3C) stan-
dardized Web Ontology Language (OWL files). Furthermore, we can capitalize on ontologies being de-
veloped for other functional genomics platforms, for example, the Functional Genomics Ontology (FuGO)
(Whetzel et al., this issue). We are working with the FuGO development group to build on and extend as-
pects of FuGO as an upper ontology for FCM development purposes.
Also, we have already developed the first draft of MI-FACE, including a detailed specification of the
gating process, its documentation, and corresponding standards for XML-based technologies. The stan-
SPIDLEN ET AL.
212
FIG. 2. Data standards for flow cytometry, project methodology. Figure shows the stepwise methodology selected to
achieve projects goals. The basic corner stones are being developed first, for example, a guideline that outlines the min-
imum information required for a FCM experiment, encapsulated within a data-centric modeling language, standardized
platform-independent file formats, standardized database representation, and a controlled terminology system. Software
tools corresponding to these standards are being created in order to provide reference implementations. Finally, a col-
laborative web space and experiments’ compendiums will support reproducible research analysis, including possibili-
ties of verification by independent researchers.
6204_19_p209-214 7/24/06 1:54 PM Page 212

dardization of the gating process is presently our most well-developed component, reflecting the high need
for standardized gating procedures. The lack of a shared representation of gates in FCM prevents a variety
of collaborative opportunities to recreate experimental methods and results. We have developed a detailed
description of the gating specification, a W3C Schema usable to validate gating XML files, user docu-
mentation, and a set of examples of gating XML files. The first version of a reference platform indepen-
dent software tool is being developed (named FACE-Java) which can read an FCS data file, along with an
accompanying XML file describing gates according to the specification we have developed, and process
this information to provide descriptive statistics. The release version of the FACE-Java software package
will be useful to validate compliance of alternative software tools implementing the gating standard.
However, Java is not the only platform that we are focused on. In order to support statistical analysis in
FCM an R package (RFlowCyt) is under development. The R Project for Statistical Computing is a very
popular open-source research platform for evaluating and implementing statistical methods. RFlowCyt pro-
vides a platform for rapid prototyping of statistical methods, report generation, and as a sophisticated cross-
platform (operating systems) data analysis tool. It currently can import data from FCS 2.0 and 3.0 files,
provides a preliminary interface for gating, and computes post-gating distributional tests for two sample
comparisons.
CHALLENGES
When any data standard is developed, the biggest challenge is insuring acceptance in the wider com-
munity. To cover the widespread acceptance of our work, it is critical that development of the standards
takes place in a open and collaborative manner that involves the entire FCM community. Therefore, we
are collaborating with the international standards body for FCM (ISAC), which will be solicited for input
and approval as the standards mature. We are also involving biologists from various fields of FCM ap-
plication, bioinformaticists, and members of the FCM software and hardware industry. Moreover, we be-
lieve that providing further motivation besides being standard-compliant can increase the likelihood of the
acceptance. We are therefore developing a software tool that enables and fosters the exchange, re-explo-
ration and re-interpretation of data and analyses by scientists. The fundamental tenet of scientific research
is that the published results of any study have to be open to independent validation or refutation (Quack-
enbush, 2004). With journal publications authors are trapped within a format and a language that is not
conductive to the complete description of software manipulations performed on data. The audience is sep-
arated from the actions and details of the algorithms used and are often forced to make assumptions re-
garding computational details which can result in completely different results. Our compendium will not
only be a traditional comprehensive compilation of a body of knowledge and experiment results, but also
it will contain a kind of active experiment result document (Gentleman et al., 2003). Such an active doc-
ument will be linked to the FCM raw data (FCS files) and experiment and analysis details including gat-
ing descriptions, and it will be created automatically and dynamically based on the linked information.
This feature not only motivates researchers but it also significantly facilitates reproducible and extensible
FCM data analyses.
CONCLUSION
In short, as FCM-based analyses expand in their size and complexity, there is now a critical need for the
development of high quality FCM data standards and tools to facilitate FCM-based research. Through the
standards we develop, along with associated tools that foster the exchange and re-exploration of experi-
ments by scientists, we hope to significantly accelerate more sophisticated and collaborative research in-
volving FCM data. This work has the potential to impact on diverse research fields, from manipulation of
stem cells, to microbiological analyses of our environment.
To obtain current information about this project, to download the proposed specifications, or to join the
discussion concerning the standards under development, please visit our web site (www.flowcyt.org).
DATA STANDARDS FOR FLOW CYTOMETRY
213
6204_19_p209-214 7/24/06 1:54 PM Page 213

SPIDLEN ET AL.
214
ACKNOWLEDGMENTS
We thank Dr. Ger van den Engh for kindly providing the Sargasso Sea sampling figure. R.R.B. is sup-
ported by the Michael Smith Foundation for Health Research. The project is supported by the grant num-
ber NIH/NIBIB R01 EB-5034.
REFERENCES
BALL, C.A., and BRAZMA, A. (2006). MGED standards: work in progress. OMICS (this issue).
BODDY, L., WILKINS, M.F., and MORRIS, C.W. (2001). Pattern recognition in flow cytometry. Cytometry 44,
195–209.
BONETTA, L. (2005). Flow cytometry smaller and better. Nat Methods 2, 785–795.
BRAYLAN, R.C. (2004). Impact of flow cytometry on the diagnosis and characterization of lymphomas, chronic lym-
phoproliferative disorders and plasma cell neoplasias. Cytometry A 58, 57–61.
BRAZMA, A. (2001). On the importance of standardisation in life sciences. Bioinformatics 17, 113–114.
BRAZMA, A., HINGAMP, P., QUACKENBUSH, J., et al. (2001). Minimum information about a microarray experi-
ment (MIAME)—toward standards for microarray data. Nat Genet 29, 365–371.
CHICUREL, M. (2002). Bioinformatics: bringing it all together. Nature 419, 751, 753, 755 passim.
DE ROSA, S.C., BRENCHLEY, J.M., and ROEDERER, M. (2003). Beyond six colors: a new era in flow cytometry.
Nat Med 9, 112–117.
FIEHN, O., KRISTAL, B., VAN OMMEN, B., et al. (2006). Establishing reporting standards for metabolomic and
metabonomic studies: a call for participation. OMICS (this issue).
GASPARETTO, M., GENTRY, T., SEBTI, S., et al. (2004). Identification of compounds that enhance the anti-lym-
phoma activity of rituximab using flow cytometric high-content screening. J Immunol Methods 292, 59–71.
GENTLEMAN, R., and LANG, D.T. (2003). Statistical analyses and reproducible research. Available at: www.bio-
stat.harvard.edu/~rgentlem/Pdf/RR.pdf.
HENGEL, R.L., and NICHOLSON, J.K. (2001). An update on the use of flow cytometry in HIV infection and AIDS.
Clin Lab Med 21, 841–856.
HERZENBERG, L.A., PARKS, D., SAHAF, B., et al. (2002). The history and future of the fluorescence activated cell
sorter and flow cytometry: a view from Stanford. Clin Chem 48, 1819–1827.
LEIF, R.C., LEIF, S.B., and LEIF, S.H. (2003). Cytometry ML, an XML format based on DICOM and FCS for ana-
lytical cytology data. Cytometry A 54, 56–65.
LE NOVERE, N., FINNEY, A., HUCKA, M., et al. (2005). Minimum information requested in the annotation of bio-
chemical models (MIRIAM). Nat Biotechnol 23, 1509–1515.
LOMAS, M. (2004). Taking a closer look at the ocean. Bermuda Biological Station for Research, Inc., annual report.
Available at: www.bbsr.edu/ar04.pdf.
ORCHARD, S., MONTECCHI-PALAZZI, L., HERMJAKOB, H., et al. (2005). The use of common ontologies and
controlled vocabularies to enable data exchange and deposition for complex proteomic experiments. Pac Symp Bio-
comput 10, 186–196.
QUACKENBUSH, J. (2004). Data standards for “omic” science. Nat Biotechnol 22, 613–614.
SEAMER, L.C., BAGWELL, C.B., BARDEN, L., et al. (1997). Proposed new data file standard for flow cytometry,
version FCS 3.0. Cytometry 28, 118–122.
TAYLOR, C.F., HERMJAKOB, H., JULIAN, JR., R.K., et al. (2006). The work of the Human Proteome Organisa-
tion’s Proteomics Standards Initiative (HUPO PSI). OMICS (this issue).
WHETZEL, P.L., BRINKMAN, R.R., CAUSTON, H., et al. (2006). Development of FuGO: an ontology for functional
genomics investigations. OMICS (this issue).
Address reprint requests to:
Dr. Ryan R. Brinkman
Terry Fox Laboratory
British Columbia Cancer Research Center
675 West 10th Ave.
Vancouver, BC, V5Z 1L3 Canada
E-mail: rbrinkman@bccrc.ca
6204_19_p209-214 7/24/06 1:54 PM Page 214