Release the brakes.
Without fuel it still won't move.
Immunology was long a science of signals. But a signal changes nothing if the fuel and building blocks are gone.
That is exactly the layer the paper's mechanism sits in — DDX54 → NF-κB → CD38 → NAD⁺ depletion → adenosine → T cell suppression. CD38 is not a checkpoint molecule but an enzyme: it does not switch a signal off, it rewrites the environment.
These twenty lessons follow that chain across 37,585 cells. You check whether adenosine-generating capacity really falls with treatment — and, while measuring the metabolism of exhausted T cells, discover that your own 'exhaustion score' was an activation score all along.
How you'll learn
We bind CD38 → NAD⁺ → adenosine into one axis and measure it across four conditions. Adenosine-generating capacity falls steadily, 0.892 → 0.580 — the same direction as the paper's Fig. 7G.
In this dataset the top Cd38 cell type is the B cell — CD38 is a classic B-cell marker too. On a cancer cell it is a weapon; on an immune cell it is an activation marker.
The exhaustion axis correlated positively with the function axis — because PD-1 and LAG-3 are also activation markers. Our 'exhaustion score' was an activation score.
A transcriptome only shows whether an enzyme is switched on. We never measured NAD⁺ or adenosine concentration. So we phrase every sentence that way — for all twenty lessons.