[Cytometry] Throwing "logs" on the fire

McCoy, J. Philip (NIH/NHLBI) [E] mccoyj at nhlbi.nih.gov
Wed May 18 11:41:45 EDT 2011

>We've always felt that log SS is a much better way of viewing PBMC or BAL or any mixture of lymphocytes, monos, and neuts, because it gives nice distributions.

I have to weigh in on this as I have never quite understood this. It is our experience that using log for SSC makes highly granulated NK cells, or activated T cells much harder to visualize than when using linear scaling in a FSC vs SSC plot. It also is more difficult to discern a break between monos and lymphs. How does compressing these values into a log scale make viewing these demarcations easier, unless log SSC is used in conjunction with CD45 instead of FSC for scatter gating?

C'mon, Mario, educate me.



J. Philip McCoy, Jr., PhD

Senior Scientist
Head, Flow Cytometry Core
National Heart, Lung, and Blood Institute
Associate Director, Trans-NIH Center for Human Immunology
10 Center Dr, MSC 1357
Bldg 10, Rm 8C103D
National Institutes of Health
Bethesda, MD 20892

tel. (301) 451-8824

-----Original Message-----
From: Mario Roederer [mailto:roederer at drmr.com] 
Sent: Wednesday, May 18, 2011 11:04 AM
To: thomas delohery
Cc: Purdue mailing list
Subject: [Cytometry] Throwing "logs" on the fire

Tom, you've known me far too long and too well to risk asking me to be nice :).  Note -- I changed the subject line, since it had nothing to do with the topic.

I have to take issue with what you state.  Just because the autofluorescence runs along 3-4 decades doesn't mean that when you add fluorescence to the system it will extend the range beyond 4-5 decades.

Look at it this way:  Let's assume a lymphocyte (with autofluorescence at 10 units, or first decade) expresses HLA-DR such that it's fluorescence shows up at 10,000 units (or the 4th decade).  That's a bright expression pattern already!

Now let's take an alveolar macrophage (with autofluorescence at the fourth decade, i.e., 10,000 units), and assume it expresses HLA-DR to the same extent as the lymphocyte (same number of HLA-DR per cell).  Where does it show up?  Not at the 7th decade -- rather, it's STILL in the 4th decade.  The fluorescence is now 10,000 (AF) + 10,000 (DR) = 20,000, which is only 1/3rd of a decade above the start of the 4th decade.

OK, so now you tell me that the macrophages express much more DR than lymphocytes.  Fair enough.  Do you think they express 10x as much?  If so, they appear in the 5th decade -- still only 4 decades above the unstained lymphocytes.  And note -- they appear in the 5th decade not because of the autofluorescence, but because they express that much receptor!  The AF would be only 10% of their signal!

My point is that it's wrong to think that because the autofluorescence ranges over 3-4 decades, that the fluorescence signal will range over a much greater range.  The range of fluorescence signal is NOT affected by autofluorescence!  (Think it through before you disagree with this).

If you really have a marker that is expressed so brightly as to appear 5 or more decades above lymphocyte autofluorescence, then perhaps you should use a dimmer color (like FITC) to detect it.  No need to waste PE on that reagent.  And if you do have such a marker (I'd love to know about it!!), then I'd ask how important is it to have the lymphocyte autofluorescence wasting a decade at the bottom of the scale?  Why not just put the positive marker on-scale, and if you have to make the lymphocytes reasonably squashed near the bottom -- you'll still be able to detect dim fluorescences on them anyway.

Bottom line -- autofluorescence doesn't impact need for more "logs".  Only if you need to discriminate 5+ decades of true expression level would you need more logs, and no one has ever shown me an example where that is the case.

As for side scatter -- we have been putting this on a log scale since the beginning of FACS.  (Yes, this is how I was trained by the Herzenbergs, who truly are the beginning of FACS!).  We've always felt that log SS is a much better way of viewing PBMC or BAL or any mixture of lymphocytes, monos, and neuts, because it gives nice distributions.


On May 17, 2011, at 9:38 PM, thomas delohery wrote:

> okay you guys, 
> In recent months I've been working with bronchoalveolar lavage (BAL) samples and the autofluorescence of the various immunocytes   in these samples can extend over 4-decades; i.e., unstained lymphocytes in the first decade and macrophages overlapping the 3rd and 4th decades along with some other interesting populations in between.  Once you add fluorescent markers to the mix it is virtually impossible to get all the populations on a 4+ log scale.  HLA/DR is particularly problematic, esp. when you are attempting to identify subpopulations of dendritic cells other APCs as well as activated T cells.  Running the samples twice after changing PMT voltages, redoing compensation, etc., is something I really want to avoid as it makes the analysis less incredibly difficult, not to mention the analysis.  I think I need more logs.  
> Also, how about a 1-log scale for 90' side scatter?  that would be very useful.
> now be nice, TomD

Cytometry mailing list
Cytometry at lists.purdue.edu
Search the list archive at  http://tinyurl.com/cytometry

More information about the Cytometry mailing list