[Cytometry] Logs on the fire

thomas delohery thomas.delohery at verizon.net
Wed May 18 22:26:46 EDT 2011


Howard, 
1) yes, point taken, it's not the number of available logs so much as the dynamic range of the detector - something you pointed out in your original post.  
2) I didn't proofread before posting.  meant to say:  Running the samples twice after changing PMT voltages, redoing compensation, etc., is something I really want to avoid as it complicates and more than doubles the data acquisition time, not to mention the analysis.  

TomD

On May 17, 2011, at 11:50 PM, Howard Shapiro wrote:

> Tom 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.  
> 
> Agreed. Now, if the bottom of your 4-decade scale is equivalent to 100 molecules of label, and your autofluorescence is into the 4th decade, it is equivalent to hundreds of thousands of molecules. But suppose you were looking at a DNA dye in that channel, and had dye bound to only 5% of the bases in each nucleus. That would give you hundreds of millions of dye molecules, obviously swamping the autofluorescence, and you'd theoretically need a 6-decade log scale. OK, now consider the voltages with which you're dealing. These days, input voltage ranges for A-to-D converters are typically no more than 2 volts, and that represents the highest amplitude signal coming out of your PMT preamps. The bottom end of a 4-decade range is 200 microvolts; the bottom end of a 6-decade range is 2 microvolts. Let's suppose your preamp puts out 100,000 volts per ampere in, a not atypical figure. If you do the math (Table 4-4 on p.203 of PFC4 goes into this) you will discover that when you try to keep the biggest signal on scale, you may have the low end down below the few photoelectrons you need from the PMT cathode to produce a detectable signal if you try to put signals from a single detector on a 6-decade scale. You'd be best off running two 4- or 5-decade detectors at different gains if you really needed the 6-decade dynamic range. Remember, what I said about 4 or 5 decades representing the useful range was intended to be restricted to immunofluorescence and similar measurements.
> 
>> 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.  
> 
> Say what? (especially that second line)
> 
>> I think I need more logs.  
> 
> Maybe, but see above.
> 
>> Also, how about a 1-log scale for 90' side scatter?  that would be very useful.
> 
> In some circumstances, yes. If you're not using log amps, it's fairly easy to display different types of data on scales with different numbers of decades.
> 
> Now, on side scatter, the marine biologists would like to display data from viruses, bacteria, various size ranges of plankton, etc., but, although there are still plenty of at least some fish in the sea, the only FISH that gets into flow cytometers is done on smaller objects, and whales, which might require decades of decades, are right out. 
> 
> With regard to fluorescence, since much of their subject material comes preloaded with PE, APC, etc., marine biologists have come to realize that the spectra of any labels they need to use had best not overlap completely with those of the molecules nature has already built in.
> 
> Einstein supposedly said that "Everything should be as simple as possible, but no simpler." Most of us in flow cytometry have become aware that what, for us, is as simple as possible is often not simple at all. 
> 
> -Howard
> 
>> 
>> now be nice, TomD
>> 
>> 
>> On May 9, 2011, at 11:15 PM, Howard Shapiro wrote:
>> 
>>> Mario wrote, in response to Bill Eades:
>>> 
>>>> I do completely agree with you on the issue of dynamic range.  Every vendor is striving to "one-up" the next.  Claims of "7 logs of dynamic range" or whatever ... these are hogwash.  I don't care if the electronics IS capable of more than 5 logs of dynamic range (so please, vendor salespeople, do not email me claiming that your cytometer is better than others for this reason).  Biologically, 4-5 decades is the maximum useful range.  Most detectors have this as a useful range.  Just because your instrument goes to 11 doesn't make it better than one which goes to 10.  And... incidentally... let's remember that all the numbers on our axes, all the MFI values .... the magnitude of these numbers is entirely meaningless.  Doesn't matter.  No information.  (Please, don't email me telling me that your company's cytometer's values are relevant.  They're not.)
>>> 
>>> Mario is absolutely right; although one can make electronics that will respond over a six- or seven-decade dynamic range, the biology with which we typically deal is what limits the effective dynamic range of our measurements. If the top end of a four-decade log scale represents 1,000,000 MESF or 1,000,000 antibody binding sites, the bottom end represents 100 MESF or 100 antibody binding sites. Although modern flow cytometers may be capable of detecting 100 MESF when measuring a single label that is the only one excited in a given beam, most multilabel staining protocols generate backgrounds of at least several thousand MESF, and relatively few of the antigens measured in most experiments are abundant enough to bind 1,000,000 molecules of antibody. It doesn't seem that we should keep having to make this point, but we apparently do. And to think that I once expected multiparameter flow results would be routinely reported in terms of molecules per cell...
>>> 
>>> -Howard
>>> 
>>> 
>>> _______________________________________________
>>> Cytometry mailing list
>>> Cytometry at lists.purdue.edu
>>> https://lists.purdue.edu/mailman/listinfo/cytometry
>>> Search the list archive at  http://tinyurl.com/cytometry
>> 
>> 
> 




More information about the Cytometry mailing list