ting

ting clusters large-scale T cell receptor repertoire (TCRR) sequences by antigen specificity to identify antigen-associated CDR3β sequences and clusters.

Key Features:

• Efficiency: Reduces computational time compared to GLIPH, clustering large datasets in approximately one hour versus GLIPH's days-to-weeks runtime.
• Scalability: Has been applied to 26 real datasets containing up to 62,000 unique CDR3β sequences.
• Accuracy and Specificity: Selects fewer sequences in naïve repertoires with minimal antigen-specific CDR3 sequences or clusters, improving specificity relative to GLIPH.

Scientific Applications:

• T cell receptor repertoire analysis: Identification of antigen-specific CDR3β sequences and clusters to study immune responses.
• Large-scale immunology studies: Processing extensive TCRR datasets to support cohort-level or high-throughput analyses.

Methodology:

Implements a novel clustering algorithm implemented in Python leveraging numpy and NetworkX.