5 Flow Cytometry Strategies That Sun Tzu Taught Me
Sun Tzu was a Chinese general and philosopher. His most famous writing is ‘The Art of War’, and has been studied by generals and CEOs, to glean ideas and strategies to help their missions. I was recently rereading this work and thought to myself if any of Sun Tzu’s lessons could apply to flow cytometry. So I have identified 5 points that I think lend themselves to thinking about flow cytometry.
“Quickness is the essence of the war.”
In flow cytometry, speed is of the essence. The longer the cells are out of their natural environment, the less happy they are. If you are performing analysis, after isolation, staining time will be dependent on what the target is. Fortunately, for analysis, it is possible to fix the cells and store them till your time on the cytometer. Of course, care must be taken when fixing cells, as fixatives can impact the quality of the fluorochromes.
Speed is especially important in cell sorting experiments. The table below shows some calculations for how long a given sort would take based on some assumptions. Including the number of events needed for the downstream experiment and the frequency of the target cell.
Now the question arises, how to decrease this time? The answer to that question is to look into getting rid of some of the non-target cells using a depletion technique like magnetic beads. Take the example of 100 million cells with a target cell with a frequency of 0.1%, and needing 100K cells for your downstream application. If we sort at 20K events per second, it would take about 83 minutes to sort. However, if we deplete those cells we don’t need (90%), which would take about 30 minutes, we would be down to only 10 million cells, which we can sort in about 10 minutes. Thus the depletion saves us about 40 minutes.
“Plan for what is difficult while it is easy, do what is great while it is small.”
At the beginning of the experimental design phase, one thing that should be developed is a statistical analysis plan. By establishing this at the beginning of the experiment, you can avoid common errors such as p-hacking and HARKing, among others.
More importantly, it also helps to determine the size of the experiment based on the power of the experiment. It is important to remember that you need to plan your experiment based on the amount of resources and money you have. This may mean you cannot answer the original question. For example, trying to prove a 5% difference requires 30 samples of each experiment and control. But you only have funds for 15 samples each. Means you can only prove a 10 or 15% difference.
The other thing to consider is your threshold, the ⍺ value. This is a measure of the change of committing a Type 1 error. While setting this value, it is worth considering the consequences of that false positive. This handy chart can help you decide. Don’t rely on the generally accepted standard of 0.05. Rather make sure your threshold is set based on the experimental question.
“If the enemy leaves a door open, you must rush in.”
We have a plan when we start our experiments. We’re trying to discover some new facts about the cells we are studying. However, sometimes the cells may want to show us something else. I know I am ascribing actions to cells that are not true. But we need to be open to the idea that we may find something unexpected and interesting in our experiments. Especially with high-dimensional experiments, we have opportunities to identify populations previously not observed. Jonathan Irish called these populations ‘cyto incognito’ as they have been overlooked in more traditional analysis.
Another way to rush in is to take full advantage of both automated analytical tools and multi ‘omics techniques. Either way, be nimble and ready.
“Opportunities multiply as they are seized.”
Flow cytometry is a powerful and evolving technology. To be successful in using flow cytometry, it is important to learn about the best practices in the design and execution of your experiments. Sure, your lab may have a coffee-stained old notebook from the PI’s time in the lab, and everyone goes to that to find the protocols that the lab uses. However, are those still the best practices? For example, when we had 3- and 4- color analog cytometers, we would perform some form of manual compensation. Unfortunately because of how these systems worked and processed data, it was very easy to overcompensate the data. Many experiments were saved because common colors used – FITC, PE, PerCP and APC, played relatively nice with each other, spectrally speaking.
Nowadays, with numerous parameters being the norm, it’s critical to perform automated compensation following the best practices, such as discussed in this blog. With the proliferation of full spectrum cytometers, automated unmixing is critical.
Another practice that needs to be consigned to the dustbins of history is the Isotype control. This blog discusses the issue in greater detail. In general, the isotype control can only show if blocking was successful and should not be used to set positivity.
There are many opportunities for further education, and you should take full advantage of what is available. Your experiments and data will thank you .
“To know your enemy, you must become your enemy.”
This quote may seem as an odd one to include, but what is the ‘enemy’ of flow cytometry? That is bad data, a failed experiment. These enemies represent lost opportunities and wasted time and money. To determine how to prevent the failure, understand where these failures occur and how to avoid the trap.
Planning is the key to knowing your enemy.
This starts with knowing the tools we can use to prevent our enemy from ruining our experiments. That is the role of our controls. From quality controls to compensation controls, and everything in between can help you understand where the failure popped up.
The assay development stage is the time to explore what controls are critical for you to identify your target populations. It also helps understand the expected ranges that these populations may have. Ensure to have or inspect the quality control work being done by the people overseeing the instrument. That will let you know the machine is behaving appropriately. In fact, don’t be afraid to add your own QC step into your experiment and track that as well.
Next, we need compensation controls to properly address spectral spillover. Ensure that we can identify our populations in the presence of all the other fluorochromes. Don’t forget the FMO control as well. It is useful in helping to set the negative/positive boundary.
Another useful control is a reference control. This is a standard sample that you run with each experiment that behaves in a predictable way in your assay.
One other trick I like to do when I am running a polychromatic panel is to have a plot of time vs fluorescence for each laser. Since a blockage on the waste side of the instrument can cause a slowdown of the flow. Prevent the cells from being interrogated by the appropriate laser at the correct time. So you will see a loss of signal as shown below.
Reading ‘The Art of War’, I was struck by some of Sun Tzu’s advice, military genius by all accounts. This also applies to the science of flow cytometry. In this blog I’ve tried to share with you some of this advice. We will see it applies to flow cytometry experimental design and execution. Perhaps some may see it as a stretch. But to me, anything that helps me remember and take pause as I perform my experiments is a good thing. The statements I’ve shared remind me to be alert as I go about my research efforts. Let me leave you with one bonus quote that I think doesn’t require much explanation. “You have to believe in yourself.”
To learn more about important control measures for your flow cytometry lab, and to get access to all of our advanced materials including 20 training videos, presentations, workbooks, and private group membership, get on the Flow Cytometry Mastery Class wait list.