A Cell Sorter in Every Lab. Can Core Facilities Survive?

 It happens in every industry across all times - What initially requires sophistication and expertise becomes simpler and more accessible to the masses and the former "experts" feel threatened and rail against the advancements.  For many years, Flow Cytometry core facilities cornered the market on ALL cytometry taking place at an institution.  As instruments became easier to use, some facilities allowed their users to begin operating the analyzers unassisted.  Affordable flow cytometers first came to market in the form of the Guava with some success, and then, in late 2006, Accuri exploded on the scene with it's affordable, easy-to-use C6 analyzer.  Undoubtedly, these instruments were marketed to individual investigators seemingly bypassing core facilities altogether.

The last stronghold of core facilities seemed to be cell sorting.  Alas, these instruments are sufficiently complicated as to assure even the most skittish of core facility technologist. That was, until easy-to-use sorters became more available.  The FACSAria (BD) was marketed as the first bench-top cell sorter capable of doing everything its more complicated predecessors could do.  That didn't pan out so well, much to the chagrin of BD.  However, after multiple iterations of the FACSAria, as well as other, easy-to-use cell sorters, we're on the cusp of a turning point in cell sorting, much like that day in 2006 when Accuri launched its C6.  In fact, here at the University of Chicago, we've jumped on this bandwagon with both feet, adding the BioRad S3 Cell Sorter to our group of cell sorters.  In addition to training users to operate the FACSAria's, we now can spend much less time training users to sort on the S3.  For a user who is familiar with the general operation of a flow cytometer, we can get them proficient on the S3 in less than a half hour.  Now, roughly one third of simple sort clogging up the FACSAria schedules can be done on the S3 with no increase in facility personnel and minimal increase in facility operating expenses.  Win/Win/Win!

However, some of us in the field may feel like this is yet another assault on our job security.  "If I don't control the sorting, what will I do all day."  I think this view is extremely myopic.  From an economic standpoint, it's always better to get something for nothing than to have to put real resources into doing it.  When you think about it more closely, if we don't have to expend resources into operating cell sorters, we can focus those resources on other important things.  I mean, it's sort of silly to have a trained professional sitting in front of a cell sorter drawing regions around GFP+ cells.  These types of sorts can easily be transferred back to the user once she is trained to operate a basic cell sorter.  This frees up time for the experts to focus on the cutting-edge stuff or development work.

This whole discussion then turns to a more philosophical discourse on what the role of resource core facilities will be in the future.  It's time to pivot, folks!  Gone are the days of a facility with 1.25 FTE's per cell sorter, staring at dots popping up on the screen all afternoon.  The business side of me wants to focus less on hardware recharge and more on services.  Why be a warehouse of hardware, when you can be the (more lucrative) service center.  Let's assume, for a moment, that cell sorters like the S3 or Sony's SH800 become so solid and affordable that many labs decide to buy their own.  Facilities focused solely on hardware recharge revenue will quickly spiral towards obsolescence. Facilities focused on services will already have other revenue streams to compensate for the lost hardware recharge.  Allow me to illustrate with a few examples.

Sample processing as a service (SPaaS):  An investigator has an idea for an experiment on a cohort of patients seen in the clinic.  We, as a core facility can coordinate pick-up of blood tubes from the phlebotomist, process the blood, bank the plasma and cells, and then stain the cells using standard panels and perform cytokine bead assays, microparticle analysis, or other assay on the plasma.  Translational applications like this are not only becoming more and more common, but many times those investigators in the best position to do these studies do not have large research labs set up to do this themselves.  Since we're not sitting in front of the FACSAria 8 hours a day, we now have time to develop these panels, and market this service.

Internal Instrument Service and Maintenance (IISM):  Investigators start purchasing their own analyzers/sorters, however, they may not have the skill, time, or resources to perform routine service and maintenance.  We, in the core facility already have the know-how to fix many of the issues on these instruments, and we have the SOPs in place to perform the necessary maintenance and quality assurance.  Instead of paying a service contract fee to BD AND having to do all the routine maintenance themselves, the investigators could contract the flow core facility to perform maintenance and service.  The core facility can, in turn, take out a self-insurance policy to maintain the instrument and recover actual costs.  The more instruments the facility can perform this service on, the better insulated you can be against catastrophic incidents.  This is essentially the model I've been using in my own lab; Budget 50% of the cost of a service contract, and pay for service calls as needed.  In the 15+ years I've been around, we've never lost money.  This works only because the number of instruments is sufficiently large.  The use, operation, and initial investment of the hardware took place completely outside the core facility.  The investigators' staff and students need training?  We can do that!  They want to assure the instrument is performing optimally?  We can do that!  They need someone to come fix the instrument because it's not working?  We can do that!  Or, if we can't, we call the OEM and pay for a service call.

Of course, there will still be other technologies associated with a traditional core facility that will still follow the normal hardware use recharge model.  High-end cytometers/sorters, imaging cytometers (a la the ImageStream), mass cytometry (CyTOF), etc...  However, we can anticipate these technologies following suit.  Soon, the CyTOF will be super easy to run, and imaging technologies will be simpler and simpler (e.g. the FlowSight).  As these technologies become commoditized like analyzers and now cell sorters, it's going to be service-focused facilities that will stand the test of time.

ABRF 2013 Recap - A flow cytometrist's take.

On the ABRF about page, it states, "The Association of Biomolecular Resource Facilities is an international society dedicated to advancing core and research biotechnology laboratories through research, communication, and education."  But, if you know anything about ABRF, you're likely involved in a molecular biology-based core facility, namely a genomics, proteomics, or related facility.  I guess I always knew about ABRF, their meetings, and what they were about, but somehow it never really interested me much.  I mean, anything with the word "biomolecular" in it gives me not-so-pleasant flashbacks to biochemistry classes.  So you can see why a guy who spends all his time working with whole cells might not take a second look at marketing materials from ABRF.  However, it seems as if the tide has shifted.

What piqued my interest this year was some interesting movement in two of the ABRF research groups that had formed in recent years.  You see, unlike other societies which may only focus on annual meetings, ABRF has interest groups that form with the intent of doing research projects.  A core group of ABRF members with common interests (e.g. flow cytometry) may come together and propose research projects to work on.  ABRF supports these efforts by providing the necessary sponsorship. A newly formed Flow Cytometry Research Group (FCRG), and a recently revived Antibody Technology Research Group (ARG) were working on some projects that seemed really interesting and very pertinent to what I do.  Seeing as I now share my time between our antibody production and flow cytometry cores you can probably guess why I'm excited by these two research groups.  Combining this with the general core facility management stuff that's always happened at ABRF pretty much made up my mind about attending this year...and I'm glad I did!

Sure there were some interesting talks about exome sequencing and insanely parallel westerns, and even the need to foster convergent technologies in order to make inroads into cancer research, but the real highlights came at the FCRG and ARG meetings.  The ARG group had been working on a modified immunization strategy to both increase the initial immune response as well as prolong that response in order to trick the immune system into making antigen specific, antibody secreting B cells.  One might not think of cutting edge technology when talking about novel monoclonal antibody production, but they had some interesting ideas.  For example, using CpG's in combination with a standard adjuvant (e.g. Freunds) when an antigen doesn't seem to be eliciting a good response.  This would be hugely important information for our Antibody facility.  The FCRG also had some interesting data surrounding the ill effects on cell function after cell sorting.  Everyone has their anecdotes about sorting at high pressure vs. low pressure, or on a jet-in-air sorter vs. cuvette sorter, but there's not much data out there in a well-controlled experiment.  Again, hugely important information.  

2013 ABRF President, David Friedman presenting Lee & Len Herzenberg with the ABRF Award

Another interesting draw for me was the fact that Drs. Lee and Len Herzenberg were being honored with the ABRF Award, and anyone hanging around with any involvement with flow cytometry got to get in a group picture with the Herzenbergs.  I have to admit being a little star-struck around them.  I really wanted to ask them for an autograph, but I didn't.  We were treated to some behind-the-scenes photos of the early days of flow cytometry.  And, it was also interesting to note how this development coincided with the development of the first personal computers.  In fact, Lee spoke about some of the first computer programs built by Wayne Moore and Dave Parks.  Probably the best part were the pictures of the 1970s versions of Wayne Moore and Dave Parks... Yeah, they look pretty much the same.

Member of the Flow Cytometry Research Group posing with the Herzenberg's following their tandem ABRF lecture

I have to say, I was quite impressed with the meeting in general.  The company that assists ABRF in putting on this show does a stellar job, and since I'm sort of involved in putting on shows on a much smaller scale, I definitely pick up on those things.  It definitely had the grandeur of a CYTO meeting, but you didn't feel completely lost in the crowd.  The biggest let down for me was the exhibitor area.  Not that it was poorly set up or anything like that, it's just none of my people were there.  One measly flow cytometry exhibitor was there, and only because they're brand new in flow cytometry and also do some stuff in the molecular biology area.  Thanks Bio-Rad. Other than that very minor demerit, it was a great conference, even for a flow guy.  I'll predict that the next time CYTO is in Europe, the US cytometry contingent will flock to the ABRF meeting.  You heard it here first, folks!  Actually, I heard it first from someone else at the meeting and am just shamelessly taking credit for the idea.

Long live the Listserv; Death to the Listserv.

Good ole listservs (or listserves or list serves, if you will).  A staple in the academic's toolbox of communication technologies.  But the question is, how useful are they in today's hyper-connected, socially-networked, always-on environment?  Here we'll touch on a few pros/cons as well as get you plugged in to cytometry networking hubs online.  

Why I love Listservs:

• Let's face it, we all live in our inboxes, right?  So, what better place to funnel all your networking than via e-mail?  It's fast, easy-to-use, and is easily accessible, especially with today's smart phone install base.  
• It's a pretty good networking tool.  Regarding the listserv's involvement in Cytometry, there has been no better platform to allow end-users to interact with field experts.  Of course, we cannot mention the words 'cytometry' and 'listserv' in the same breath without pointing you to the preeminent spot for networking, the Purdue Cytometry List. For years (over 20, now) the Purdue List (as it's commonly referred to) has allowed cytometry professionals to interact and network.
• Being an active Listserv participant also gains you exposure, which can lead to new opportunities.  

Why I wish Listservs would crawl up into a little ball and be subjected to a slow painful demise:

• Poor search leads to lazy researchers.  The fact that it can be difficult to search listservs inevitably results in people asking the same questions that were answered already on the list.  Long-time list participants may respond with not-so-pleasant remarks to such a query.  This sometimes even leads to a discussion on the merits of re-visiting previously answered questions, when we should be discussing the original question itself.  
• "Out-of-office" replies...need I say more?
• The "Good Samaritan Effect."  People who may know the answer to a posted question may pass by without helping because they assume someone else will help them out.   
• Too broad or too diverse.  Listservs typically cannot be broken down into categories, allowing people to focus in on a specific area of the listserv's main topic.
• No inline rich media.  Depending on the listserv, you may not be able to attach documents, pictures, movies, etc...  Try and explain your gating strategy only using words...It's not fun.
• Wikipedia tells me that LISTSERV was developed in 1986.  Nineteen Hundred and Eight Six!!!!!  Do you know how long ago that was?  It was the last time the Bears won the Superbowl (OUCH!).
• Do we really need to be shuffling around emails to 4000 people on a listserv?  Answer: No.

21st century tools for a 21st century technology

Obviously, I wouldn't bring you here unless I had some alternatives to offer.  And, you can probably already guess a few of the punchlines, but allow me to state the obvious.  SOCIAL NETWORKING.  LinkedIn, Facebook, Google+, and Twitter have demonstrated the power of social networking platforms.  Already, there is a substantive presence of cytometry on each of these platforms; some of them are actually quite fruitful.  These tools allow for:

• Sharing rich media inline with text to create better communication of ideas.
• A better sense of interactivity of the group instead of a one-to-one interaction.
• Fantastically good search tools for finding exactly the information you need
• Using email notification settings, you have the ability to interact as much or as little as you'd like.
• Speaking of email. Many of these services allow you to fully interact with the group using the email interface, if that's more your style.
• Strong sense of community - Via avatars and in-depth profiles, you're able to build better relationships with colleagues.
• Expand your interaction with people on the fringe since they're already using these networks for other (personal and professional) purposes.

So, what's out there?  

Well, if you just do some searches, you're bound to find groups online.  To point out one near and dear to me, I'll plug the fledgling Google+ Cytometry Community (for which I'm one of the moderators).  Google+ as a platform, is becoming quite full-featured in this regard, and the potential to have a very interactive online community is strong.  Of course, the fact that it's backed by powerful search and a host of integrated tools (Drive, Gmail, Picasa, etc...) makes it, de facto, a force to be reckoned with.  It is becoming more and more clear that tools like Google+ are most definitely the future, and antiquated platforms like listservs are (slowly) in decline.  My advice to you?  Get out there and start interacting.  Why not hop over to Google+ and grab yourself an account (if you don't have one yet) and then stop by the Cytometry Community to say hi!

Life Technologies Attune Cytometer... Deja Vu

It wasn't so long ago that I blogged my impressions of the newly released Attune Cytometer from Life Technologies.  Although overall, the system seemed pretty good, I did find some disheartening issues with both the hardware and software.  There's no reason why this instrument shouldn't be one of the best pieces of hardware on the market, regardless of price-point.  The entire premise of acoustic focusing should lead to excellent fluorescence resolution.  So, when Life Technologies approached me again to give the Attune another look, I obliged.  The reason for the poor performance the first go around?  Poor alignment and a very generous install specification.  With some of those issues taken care of, I was able to spend a few months with both the Blue/Red laser configured system and the Blue/Violet system.  I'll reference back to the original article in various places, and highlight some of the improvements I've seen with the systems.

Fluidics:

Much of the fluidics system has remained unchanged for the most part.  A few of the biggest issues from the prior testing have been addressed specifically point #8 where there seemed to be some fluctuation of the fluorescence signal after a re-draw of the sample syringe.  That no longer happens and is not an issue.  The other issue is the return of unused sampled once you've acquired enough events.  Now, at the end of a run, you can enable a script to return the unused portion of the sample to your tube.  I'll also mention the startup and shutdown procedures here as well.  They've really done it right on this system.  You can start and shutdown the fluidics without having to log in (useful for usage tracking in a core facility).  Also, the system turns itself off after the shutdown is complete, which is nice at the end of the day.  Lastly, their performance tracking QC module works well and is very informative, yielding Levy-Jennings-type plots for all your parameters as well as giving you an overall pass/fail (similar to CS&T in DiVa).

Software:

If you've used DiVa, you've used the Attune software.  The software was pretty snappy overall, and handled large data files (>100,000 events) without hesitation.  There's still a lag switching between tubes while you're waiting for the fluidics to finish its process.  It's long enough for you to notice it, but not so long that it becomes annoying.  It's times like this however, that I wish I would have had the 96 well loader on the system.  Also, the plot scale needs to be changed.  Right now, they're displaying a 7-log scale of which the first 2 logs are pretty much useless.  In setting up an unstained control, you end up putting your cells between the 2nd and 3rd decade anyway, so there's no reason to have empty white space at the bottom of your plots.  This only adds to a user's confusion on how to optimally set up their voltages.

There is no noticeable differences in Optics or Electronic other than the introduction of a Blue/Red laser configured instrument, which is built as a 4-2 configuration.

Data:

As far as testing goes, I focused on resolution tests I had run previously on the system, which utilizes stained capture beads.  I was working on a further development of this technique which has resulted in a quantitative metric called qNORM.  The qNORM can be defined as the minimum number of antibodies bound to a cell that can be resolved from unstained lymphocytes.  For example a qNORM value of 1000 would mean that this channel on this instrument could resolve a cell type that was stained with as few as 1000 antibodies from the unstained cells.  In testing the Attune systems, I used a few other instruments in and around the flow facility for comparison's sake. The first table below summarizes the instruments used (including the laser powers and the vintage/health of the specific instrument tested).  The second bar graph shows the qNORM values for the different instruments for 5 fluorochromes.  Please note:  lower qNORM values = better resolution.

Instruments at UCFlow tested for the qNORM characterization

qNORM values for 5 channels on the respective instruments.  Error bars are calculated from 4 separate experiments.

What we can take home as far as the Attune's performance on these tests is that it falls in line with its cohort on some of the channels (similar in FITC to a FACSCanto and MACSQuant), but does surprisingly well in other channels (as good as the Gallios or Fortessa in the PECy7 and APC Channels).  You'll notice an anomaly in the PECy7 performance of the BV Attune compared with the BR Attune.  This is due to a suboptimal filter on the BV for PECy7 (a 650 LP instead of a specific PECy7 filter such as a 780/60 BP) and a non-red sensitive PMT in that channel.  Both of these deficits can be easily fixed by swapping in a appropriate filter when using that color or using a more red-sensitive PMT.

Final Thoughts:  The units I evaluated this time around performed much better on all the tests run.  The most remarkable feature of the Attune is its ability to perform on par with the more expensive cytometers in some of the channels AND retain that level of resolution at much higher flow rates.  I said it before, and I'll say it again, "There's no reason why this instrument shouldn't be one of the best pieces of hardware on the market regardless of price-point", and if they continue to develop the platform, specifically working on the software issues highlighted above, I think it could easily take that crown.

Propel Labs Avalon: A first look.

N.B.  The Avalon cell sorter is now the S3 cell sorter distributed by Bio-Rad.  It's the same hardware/software, with a new color scheme to match the Bio-Rad colors.  You can find the product's new page here:  http://www.bio-rad.com/prd/en/US/LSR/PDP/MC3PU4E8Z/S3-Cell-Sorter

I recently had the opportunity to check out the highly anticipated Avalon cell sorter from little-known Propel Labs (Ft. Collins, CO).  Little known, that is, until this year's CYTO conference where they unveiled the new 2-laser, 4-color bench-top cell sorter.  This made such a huge splash at CYTO even competitor vendors couldn't help but check it out.  After the dust settled from the whirlwind tour, would the Avalon prove to be a true, easy-to-use, bench-top cell sorter?  Read on for the answer.

Do you remember when the original FACSAria was announced?  It was billed as so easy to use, you could train all your users to run the sorter themselves.  BD was forced to quickly retract those marketing talking points as things like "Nozzle-gate" came to light.  Since then, there have been a few attempts at "user-friendly" sorters, but all them required too much hardware manipulation by the user.  So, you could imagine I was pretty skeptical when the new Avalon was touted to be easy-to-use.  As such, my evaluation of the Avalon focused primarily around this notion that I could train any of my users, who possessed a certain level of flow cytometry know-how, to sort their cells on this instrument with minimal instruction.  I was NOT disappointed.  Secondarily, I wanted to test the basics, like sorting speed, yield, purity, and to a lesser degree fluorescence resolution.  Let's begin.

The Propel Labs Avalon Cell Sorter

First, I will say the instrument is pretty nice looking.  At 70 cm wide x 65 cm tall x 65 cm deep, it's bigger than a breadbox, but not much bigger.  And the best part is, what you see here in the image to the right is the ENTIRE instrument, no wet cart, no external power supplies, no nothing.  Everything is contained within that footprint.  Behind the larger grey door on the front houses the sheath (more on that in a sec), waste tank, and DI water tank (used for shutdowns).  And here's the kicker;  You can load a tank of 8X sheath fluid, and it dilutes your concentrate to 1X on-board using the DI water.  Now, there's even less of a chance of introducing contamination when making up your 1X PBS for your cell sorters.

To test the system's sorting ability, I ran a sample using my normal cell sorting specification.  The spec we use at UCFlow is:  Event Rate = Frequency/4, and sorting a 5% population, the efficiency should be >80% and the purity should be >98%.  This spec is mostly governed by physics and poisson statistics, so if an instrument cannot pass this spec, then there is something else going on (too high of a hard abort rate, too slow of an electronics processing system, etc...).  Below are the results (analyzed in FlowJo, of course).  As you can see, the Avalon easily handled this test.  N.B. The sample was fixed/no-wash peripheral blood stained with CD3, CD4, and CD56, which is why the background is pretty high and the separation is not spectacular.  This is also why the post-sort populations actually look better; we've essentially 'washed' the sample by diluting it in the collection media.  You'll also note here the somewhat standard droplet formation values:  30 PSI, 100um tip, 38KHz.  There are 3 modes for sorting, Purify, Enrich, and Single Cell, and in Purify or Enrich, the sort envelope is 1/2 droplets.

Sorting test using 3-color staining of fixed/nowash Peripheral Blood.

The 100mW 488nm and 561nm lasers are plenty powerful to resolve most populations.  This, combined with a careful selection of Semrock filters, and an in-house developed collection lens makes the Avalon no slouch when it comes to fluorescence resolution.  The requisite 8-peak bead data is presented below.  Use this information however you'd like, but use it with caution.

8-peak data collected by UCFlow during Demo

So, we know it can sort, and we know it can resolve fluorescence adequately, but what about user-friendliness.  Well, you've reached the best part!  There's a single power button on the system; everything else is controlled through the software.  There's a one-click startup sequence that primes the fluidics, and turns the stream on.  There's a one-click, one bead alignement QC AND Drop-Delay assessment, and voila, you're ready-to-sort.  There are a few details that occur behind the scenes here that are worth mentioning, but for the average user, they won't need to know any of this.  It just works! So, what happens during the start up is a coarse alignment of the stream to the light collection pinhole.  This is completely automated, and happens via pico-motor controllers on the x/y stage of the nozzle.  The laser lines are fixed in place.  Also, at this point, the system scans through the droplet frequency and amplitude, picks the optimal node, and optimizes the drop charge phase.  After the coarse alignment, the QC/DropDelay beads are put on and the alignment is peaked.  Lastly, it does a drop-delay test by sorting these same beads and counting the beads left in the waste trough at each adjustment in delay.  In the waste trough is a small laser which excites the beads, and a fiber cable that carries any fluorescence to one of the systems PMTs.  When the PMT fluorescence drops down to zero, the drop-delay is set.  Again, all of this happens without any user intervention.  Once set, the delay is monitored and adjusted as needed, and sorts are terminated in the case of a clog.  The software, although a bit buggy yet, had a nice familiar feel taking a page from the MS Office suite and utilizing the Ribbon feature.

Look and feel of Avalon Sorter Software

The sample tube holder and collection receptacles are temperature controlled from 4C - 37C, and completely contained for excellent aerosol management.  The sample tube holder has this really cool loading action.  There's a front and rear position; the front position is where your sample is loaded, and the rear position is where the sample line is backflushed and rinsed, inside and out.  You can see the loading and unloading action in the video below.  Here, the Avalon covers are off for easy viewing.

Sort collection is equally unique.  The system sorts 2-way, with a 5 postion tube holder (each side).  So, you could load 10 collection tubes on the system, and once they fill the tubes in position 1, it'll automagically slide the next (empty) tubes into place.  No more having to stop the sort, pull out full collection tubes, and put in new empty ones.  The only complaint here is the lack of 4-way sorting and true, 96 well plate sorting.  I say 'true' because there is actually an attachment that allows you to place two 8-well strips on the collection receptacle allowing you to sort 16 wells at a time.  It certainly slows down your high-throughput sorting, but it works.

Looking into the sort chamber on the Avalon.
5-position collection receptacle in foreground

Final thoughts.  The Avalon seems to capitalize on all the little things that so many companies got wrong in the past.  Simplifying the design, minimizing the steps required to go from 'power-up' to 'ready-to-sort', and making a seemingly solid instrument, all add up to a very attractive package.  To top things off, it comes at a bargain price, starting at just 99,950.00 US$ for a 1-laser, 2-color system.  Sure, it won't replace the need for a 5-laser FACSAria III, but it can surely supplement your sorting needs by offloading those simple 2-4 color sorts clogging up your schedule.  The only question then is, How soon can you scrounge up the funds?

My 3-step approach to gating Annexin V data appropriately

Using flow cytometry to assess cell death, and more specifically, apoptosis, is an everyday occurrence in a flow core.  Technically speaking, performing the assay couldn't be easier when utilizing numerous kits available from any number of vendors.  It's typically set up in the format, add 'x' ul of reagent A to your cells, then 'x' ul of reagent B, and analyze.  So, with such a simple assay, what could possibly go wrong?... Lots!

Let's assume there are no faults with setting up the instrument or parameter voltages and the like.  Let's also assume the assay worked; that is, control cells looked relatively live, and treated cells looked nicely dead.  The possible problem I've been coyly hinting at therefore, is in the analysis.

What's typically the first thing someone does when they analyze their flow data?  They make a forward scatter (FS) versus side scatter (SS) plot and gate on the 'live' cells.  Obviously, if you're trying to determine frequencies of live and dead cells, you wouldn't use a 'live' gate, you'd extend the gate lower on the FS parameter to include the dead cells.  However, when you do so, you could be including small debris and bits of cells, and it's not exactly fair to count fragments of cells as events.  So, herein lies the dilemma:  You'd like to count whole cells, while drawing a light scatter gate that encompasses both live and dead/dying cells. Hence, My 3-step approach to gating Annexin V data appropriately.

Step 1:  Display your Annexin V channel versus your Membrane Permeability Channel (for ease of typing, I'll assume we're using Annexin V FITC and PI) on the entire, ungated data set, and draw a region around the double negative (DN) fraction.

Step 2:  Using the DN region from Step 1, gate a FS versus SS plot on these DN events, and draw a tight region around the subset of cells with low FS, now called Debris.

Step 3:  Invert this Debris gate, creating a Not-Debris gate, and use this Not-Debris gate on the total population for your downstream Annexin V vs. PI display.

To see a screencast walkthrough of the gating process, watch the video below:

Let me break this down for you using an example.  The entire point of the 3-step exercise is to properly define an area of the FS vs. SS plot that you'll call Debris, which will not be counted when generating your downstream frequencies.  Depending on how much death you induce, and how toxic your treatment is, you could have a significant amount of Debris in your sample.  It's important to note that you're not simply throwing away these events, you're just removing them from the downstream classification as Live (Annexin V neg, PI neg) Early Apoptotic (Annexin V pos, PI neg), Late Apoptotic (Annexin V pos, PI pos), or 'Necrotic' (Annexin V neg, PI pos).  These debris events belong to none of these categories and therefore should not be included.  The harm in including these debris events is that they usually end up in the Live fraction since they neither have a nucleus nor an intact membrane.  Therefore, leaving them in inflates the Live fraction of cells, yielding potentially misleading results.  The figure below shows the respective plots and regions.

Plot A shows a FS vs. SS plot of the full data set.  Plot B shows the Annexin V and PI fluorescence on the entire data set with the double negative region drawn.  In C, we are showing a plot of FS vs. SS gated on the DN cells.  Here we can see we're left with two distinct populations.  The larger population is the true Live cells (Annexin V neg, PI neg, and large), and the other population is the debris.  I use a contour plot to show a nice outline of the region that should be drawn.  This region is then applied to the top level FS vs. SS plot (as in D), and is inverted.  This inverted Debris gate (i.e. Not-Debris) is then applied to the Annexin V vs. PI plot and quadrants are drawn to define the 4 populations present. 

Not quite convinced yet?  Well, I ran a similar sample on the ImageStreamX, gated in a similar way, but now, using the imaging power of the ISX, we're able to visually confirm what each of these populations is.  To the left, we see the analogous plots and regions as prescribed in the 3-step method, and below we can see a representative pool of images from each of the regions.  The panels labeled Debris and Live came from the DN gate drawn, and as in the flow cytometry data, the FS vs. SS plot (here presented as Bright Field Area and Side Scatter) separates into 2 distinct populations.  A larger Live cell population and a smaller Debris population.  Once the Debris is removed from the downstream classification, we are left with the Live, Early/Late Apoptotic, and Necrotic fractions.  Of special note is the fact that we are not simply gating out small events, as there are small events that do in fact belong to the Necrotic or Apoptotic groups (see especially event #2269 in the Late Apoptotic panel).  We're specifically gating out events that are small and have no fluorescence, which in this case, is a pretty good definition of debris.  

Now obviously there are a whole bunch of caveats.  This works really well for a homogeneously sized cell line, and perhaps less well for a mixed population.  Also, as was briefly mentioned above, you'll need to still factor in that % Debris metric somehow.  It could be your treatment makes either live cells or debris, and excluding the Debris from the Annexin V plot may show you have a healthy profile.  However, it's important to track % Debris in addition to the other populations when comparing your conditions.  If nothing else, though, you now have a reproducible method for gating out debris from your apoptosis analysis that could actually be explained properly in the methods section of your manuscript.

Fluorish vs. Chromocyte... READY, FIGHT!

One of the more difficult tasks people encounter when they start making multicolor flow cytometry panels is the process of finding antibodies, and knowing which fluorochromes to use.  There's the primary task of finding the specific antibodies you want to use in a few fluorochrome options (hopefully) and figuring out which ones would work best on the instruments available.  And then, there's a secondary task of optimizing the fluorochrome with the antibody in line with the adage:  Dim fluors on abundant antigen, bright flours on sparse antigen.  This secondary, but equally important task is a pretty advanced skill that is difficult to automate or build into an algorithm, although some companies, like Woodside Logic's CytoGenie, do a decent job of trying to handle this.  However, in order to get to this second step, we need to first aggregate all the information from a lot of sources as to what's available.  This step can be built into a utility, and it is precisely this step which I will discuss below.  In particular, I'm going to focus on two offerings that have come online in recent years.  One is the utility built into the Chromocyte web portal** under its 'calculate' tab, and the other is a utility developed by the parent of FlowJo called Fluorish.

After playing around with both services for a bit, I'll highlight a few areas which can serve as a good comparison of the two.  The areas I'll focus on are Catalog, Hardware Integration, and Usability.

Catalog:  First off, let's look at the catalog of antibodies available from the two sites.  As expected, there's a lot of overlap, and they both have the catalogs of all the major vendors you'd expect to see.  There were, however a few exceptions.  For example, Fluorish.com doesn't have the catalog of EMD-Millipore, GeneTex, Cytognos, antibodies-online, Gen-probe Diaclone, and IMGENEX Corp - all of whom appear on the Chromocyte site.  The Chromocyte site didn't have Miltenyi Biotec, and Sony/iCyt, both of which are present on Fluorish.com.  Whether this translates into a use-case where you're unable to find a specific antibody you're looking for or not is debatable.  I'd assume that nearly every antibody commercially available is likely to be on both sites, and if there is something not available, both sites prominently display a way to let them know so they can add it.  However, it should be noted that the Chromocyte list seems to be outrageously exhaustive; allowing you to select antibody species from Alligator to Zebrafish, and everything in between.  Another nice feature of both sites is the ability to select viability or other dyes, and fluorescent proteins so you can exclude those channels from the potential list of available fluorochromes.  So far, we'll call this a truce, but wait...there's more.  Fluorish.com, not only allows you to query all these commercial sources, but you can also add your own private stash to the database and allow those custom sources to be displayed in-line with the commercial sources.  It just so happens that I'm part of a core facility that also makes and conjugates antibodies, so this is a BIG DEAL for me.  Catalog Winner:  FLUORISH.COM 

Fluorish's Panel Wizard showing Instruments in UCFlow Core.

Hardware Integration:  Next, I'm looking for the ability to integrate information regarding your facility's hardware.  Again, both sites offer the opportunity to input and save the configuration of your specific instrument.  Chromocyte uses a method where you select pre-defined filters from a list, whereas Fluorish has you input the cut-on and cut-off wavelength for each filter.  Obviously, if your specific filter is not in the Chromocyte list, you're out of luck (although you could request your filter to be added).  Using the Fluorish method allows you to precisely input your specific/custom filter, but it takes a bit of mental rigor (at least for my brain) as you try to calculate the cut-on and cut-off of a 595/14 BP filter.  Fluorish, however, takes things one step further by allowing you to export your FACSDiVa configuration file from CS&T and drop it in, yielding a hands-free hardware configuration process.  Both services have built in a layer of groups, which basically means I am set up on their site as a core facility, and other users can be part of my group, which gives them access to all the configurations I've already set up.  As you can see, both utilities have done a very nice job allowing you to build panels specific to your hardware configuration, however with the abundance of FACSDiVa systems, the extra ability to drag and drop instrument configurations from CS&T into Fluorish puts it just over top of Chromocyte.  Hardware Integration Winner (by a hair):  FLUORISH.COM

Usability:  Lastly, but most importantly is Usability.  This is such a huge topic, and one that's likely very personal to the individual user, that I'll simply report some of the things I really like or dislike about each of the services.  When it comes time to pick which antibodies you will use, I really like the interface of Chromocyte where they lay everything out in a grid pattern.  You can easily get a sense, at-a-glance, of how things might play out.  If there are only a couple options under one of the columns, you'll know you should probably take care of that column (i.e. channel) first.  The difficult part is that when you have a bunch of antibodies you're searching for and you take care of one of them (e.g. CD45), you'd expect the CD45 options in all the other channels would be greyed out or disappear.  However, it allows you to accidentally select CD45 in multiple colors.  Also, since the lists can sometimes be long, and if you choose one of the options that are off screen (i.e. an option you needed to scroll down to select), it would be nice if the column was greyed out or somehow highlighted to show you that channel is taken care of.

Chromocyte's interface for picking conjugated antibodies

Fluorish's answer to these problems is that it dims both antibody and fluorochrome options that remain once you choose your preference.  Although Fluorish's interface isn't as pretty as Chromocytes, it's definitely more functional and easier to use.

A nice feature of Fluorish that's missing from Chromocyte is the ability to specify a specific clone of an antibody.  So, if I know that I want to use RPA-T4 instead of OKT4, I can select that specific clone to refine my search.  

An advantage of Chromocyte is the fact that it lives completely in the cloud.  There's no software to download or upgrade, and so you're always using the latest version.  With Fluorish, there are two sides of the equation.  The first is Fluorish.com, which has limited functionality, but controls things like accounts, and core facility membership, etc...  The second part is a program that you download called the Panel Wizard (currently on beta version 1.0b34).  It works well, but completely fails at one crucial part of the equation... Mobile.  Now Chromocyte is by no means optimized for mobile, but it at least works.  Mobile is such a huge segment of my daily workflow, it's tough to forgive this omission on Fluorish's part.  However, it's desktop application does work quite well, and with it's nice usability features discussed above, it more than makes up for that omission.  Usability Winner:  FLUORISH.COM  

So, there you have it.  In terms of a free utility to help you build panels for your multicolor flow cytometry needs, Fluorish.com appears to be a better option.  **Of course, it must be said that Chromocyte has A LOT more to offer than its 'Calculate' tab.  With its wide breadth of resources on flow cytometry techniques, services, news, and tutorials, Chromocyte is a one-stop-shop for the flow cytometry community.  They've even plugged one of my blog posts, so you know they have good taste!  Now we just need Fluorish and Chromocyte to get together and make a baby so we can have to best of both worlds.