PROBABILITY STATE MODELING: A NEW PARADIGM FOR CYTOMETRIC ANALYSIS - Flow Cytometry in Drug Discovery and Development - page 287

know this. The analysis system is showing this to be true, and that serves as another
confirmation that we are on the right track in modeling the B-cell lineage. However,
we can use the literature to squeeze even more information from the data. We know
that when B cells become plasma cells, they lose CD20. Thus, we can have our CD20
parameter profile drop to negative at the end to try to catch these plasma cells when
they repopulated the marrow (Figure 14.10d).
So far, we have examined five parameters. We can see all the correlations among
these five parameters in our overlay plot (Figure 14.10d, bottom plot). As discussed
earlier, we can also read off all the subpopulation percentages because the state
frequencies are uniform. We can directly estimate that TdT þ cells are around 10%,

is 42%, and CD20 þ is 27%. As shown earlier, these percentages are
accounting for measurement overlap. Thewidths of the bands shown in all the overlays
are related to the population

CD20

s confidence limits at each stage in the B-cell progression.

14.7.3 The B-Cell Scaffold

These five parameters are enough to define all the important stages of B-cell
development. They form a kind of B-cell scaffold that other parameters can be
compared against. Figure 14.11 shows how CD45, CD43, CD23, CD34, and FSC
relate to this B-cell scaffold.

FIGURE 14.11 B-cell scaffold determines the profiles of CD45, CD43, CD23, CD34, and
FSC. (See the color version of the figure in the Color Plates section.)