Back

A Donor T-Cell Receptor Structural Signature Determines Alloreactive Potential and Predicts Acute Graft-Versus-Host Disease

Wang, X. K.; Uzuni, A.; Shi, L.; Harle, D. W.; Macedo, R. J.; Pressler, M.; Gordillo, C. A.; Belay, K.; Chakrabarti, S.; Fuller, J.; Hexner, E. O.; Loren, A. W.; Porter, D. L.; Mapara, M. Y.; Sykes, M.; Azizi, E.; Reshef, R.

2026-07-11 immunology
10.64898/2026.07.08.737087 bioRxiv
Show abstract

Acute graft-versus-host disease (GVHD) remains a lethal barrier to successful allogeneic hematopoietic cell transplantation, yet pre-transplant donor selection entirely ignores hypervariable T-cell receptor (TCR) architecture. Here, by characterizing over 64,000 alloreactive clonotypes, we demonstrate that human alloreactivity is dictated by a constrained, predictable baseline structural signature. Pathogenic, tissue-infiltrating alloreactive T cells exhibit significantly shortened CDR3{beta} regions, altered antigen-facing biophysical features, biased VJ gene usage and extensive inter-donor sharing originating from public anti-pathogen memory reservoirs. These potent clones natively cluster within the high-frequency fraction of the unstimulated baseline donor repertoire. We introduce R50, an assay-independent metric quantifying this clonal dominance, which independently predicted a six-fold increased risk of acute GVHD in a cross-institutional cohort. This scalable in silico platform shifts pre-transplant risk stratification from HLA typing and demographic surrogates to precision immune-receptor modeling.

Matching journals

The top 10 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.