TCRosetta: a powerful server for analyzing and annotating T-cell receptor repertoire

Background: Analyzing the T-cell receptor (TCR) repertoire is important with the advent of precision medicine and immunotherapy, as TCR repertoire could serve as a biomarker of immune response and disease progression. Though many TCR analysis methods have been developed for analyzing various aspects of the TCR repertoire, the usage of these methods requires experience in bioinformatics and is technically cumbersome. There is an urgent need to develop an easy-used online server for TCR repertoire analysis. Methods TCR CDR3 sequence with clinical information were collected from TCRdb. In TCR repertoire general analysis, Renyi entropy and 1-Pielou’s index was used to calculate diversity and clonality of the TCR repertoire, respectively. GIANA and iGraph were used to discover possible disease-specific TCR CDR3 sequences and construct a TCR network in network analysis. OLGA was used to calculate the generation probability of a CDR3 sequence in healthy individuals in public analysis. PHATE was used to embed TCR repertoire in embedding analysis. Results In this study, we introduce TCRosetta, a powerful platform for analyzing and annotating TCR repertoire. Above 244 million complementary determining region 3 (CDR3) sequences of TCR beta chain (TRB) with disease information are integrated into the webserver, enabling big-data-based CDR3 annotation for the first time. The main functions of TCRosetta are as follows: (i) General feature analysis for TCR repertoire, including diversity, V/J gene usage, CDR3 length distribution, and clonality etc.; (ii) Annotate the disease preference of TCR repertoire and TRB CDR3 sequences; (iii) TCR repertoire network construction and analysis; (iv) Calculate generation probability of TRB CDR3 sequences. TCRosetta is the first comprehensive online server for TCR repertoire analysis and is useful for immunology research.