FAQ  •  Register  •  Login

Sensitivity of CyTOF vs. Fluorescence Cytometry

<<

maecker

Contributor

Posts: 23

Joined: Thu Oct 31, 2013 7:58 pm

Post Sat Nov 09, 2013 12:30 am

Sensitivity of CyTOF vs. Fluorescence Cytometry

We are often asked about the sensitivity of CyTOF relative to conventional flow cytometry. It is a complex question, because, like flow cytometers, the CyTOF has label bias: it detects some isotopes with greater sensitivity than others. This is a factor of the detector mass window, with maximal sensitivity in the upper-middle portion of that window (peaking around AM 168-170 on the CyTOF 1). However, the ratio of highest to lowest sensitivity isotope is only about 3-4; while in conventional flow cytometry, fluorochromes can vary more than 10-fold in brightness. Of course, there are also many different types of flow cytometers, with differing sensitivities for any given fluorochrome. Nevertheless, we can summarize the situation as follows:
    -There is no mass channel as sensitive as PE (generally the brightest fluorochrome on a conventional flow cytometer)
    -That said, there are about 40 CyTOF channels available, and all are better than the dimmest fluorochromes
    -The use of multiple markers in CyTOF can partially overcome lack of resolution of a population using any one marker (but this may require multidimensional analysis, not sequential gating of 1- or 2-dimensional plots)
    -Antibodies that are dim in fluorescence are generally dim in CyTOF, and vice versa; though there are a few discrepancies that probably have to do with the conjugation chemistry of CyTOF polymers
<<

Ofir

Master

Posts: 75

Joined: Thu Nov 07, 2013 12:46 pm

Location: US, CA

Post Tue Nov 26, 2013 7:26 pm

Re: Sensitivity of CyTOF vs. Fluorescence Cytometry

I think an important metric here is resolution - the ability to resolve signals of varying intensities. In FACS we typically use Stain Index.
In this sense we can consider 3 cases:
    (1) discrete expression or positive vs. negative - e.g., CD3e,
    (2) several distinct populations - e.g., CD45hi/dim/neg, and
    (3) continuum of expression - e.g. CD62L.

One of the biggest advantages of mass cytometry is the very low background - essentially zero from the instrument. Compared with the natural auto-fluorescence of cells in all FACS channels (to varying degrees). The background that we detect in mass cytometry is mostly from non-specific binding of the antibody, which is the same regardless of the tag (with a few notable exceptions).

So, to the point, a very low signal in CyTOF still allows excellent discrimination of negative and positive populations (case 1) and very good discrimination of distinct hi/dim/neg populations (case 2). Where the challenge lies is in the range of a continuous measurement (case 3). Obviously the bigger the range the better the resolution of continuous expression levels.
In many cases we use CyTOF data in comparative multivariate analysis such as with SPADE and viSNE. In these cases it is important that the range not change between samples (say 10-10,000). However, when we drill down and do manual gating (think FlowJo), this could be a disadvantage as calling the gate boundaries can become more difficult. I do think these are the minority of the cases though.

To summarize, my experience is that the resolution of CyTOF data is more than adequate for the purpose of scientific research. Now for clinical use, that's a whole different story, and one yet to be told.

Return to CyTOF general discussion

Who is online

Users browsing this forum: No registered users and 8 guests