Engineered T-cell receptors (TCRs) have tremendous therapeutic potential for targeted, T-cell-mediated killing of infected or malignant cells. A key hurdle in realizing this vision is understanding and manipulating the interactions of TCRs with their cognate antigens. Rules for TCR engagement have been difficult to define owing to the complexity of the system, and platforms for de novo TCR design remain aspirational. A leap forward is now provided in papers published recently in Nature by Dash et al.1 and Glanville et al.2 who show that at positions of high-antigen-contact probability, TCR variants tend to have conserved sequence motifs. Once identified, these motifs, can be used to predict the antigen recognition properties of newly sequenced TCRs
