Stratification of immunotherapy responses by adaptive immune receptor repertoires

A growing number of diseases are now treated with immunotherapies, which consist of interventions that suppress or stimulate the patient's immune system. Because individual humans express a unique repertoire of adaptive immune receptors, the efficacy of immunotherapies typically varies from person to person. Next generation sequencing of adaptive immune receptor repertoires, combined with machine learning or statistical analysis, has emerged as a sensitive means of stratifying patients based on their immune status, particularly in the fields of cancer and autoimmune disease therapy. The sensitivity and specificity of these approaches rely heavily on the methods of deriving features from each individual repertoire. Here, we review recent trends in stratification methods and highlight their limitations, including the need for data standardization and sharing.