GIANA

GIANA performs rapid alignment and clustering of up to 10^7 T-cell receptor (TCR) hypervariable CDR3 amino acid sequences by encoding sequences with a geometric isometry framework for large-scale immune repertoire analysis.

Key Features:

• Computational Efficiency: Achieves sequence alignment and clustering at speeds reported to be approximately 600 times faster than methods such as TCRdist while maintaining specificity.
• Isometric Encoding: Encodes amino acid sequences using a geometric isometry-based mathematical representation to preserve sequence similarity relationships.
• Rapid Query Capability: Enables rapid querying of extensive reference cohorts within minutes for large-scale comparative analyses.

Scientific Applications:

• Identification of Shared Antigen Specificity: Clusters TCR sequences to identify shared antigen specificities among receptors.
• Disease-Specific Receptor Discovery: Facilitates identification of candidate disease-specific TCRs from large-scale repertoire datasets.
• Multi-Disease Diagnostic Comparison: Allows comparison of unseen TCR-seq samples against reference cohorts to differentiate samples across cohorts associated with cancer, infectious diseases, and autoimmune conditions.

Methodology:

GIANA transforms amino acid CDR3 sequences into an isometric geometric encoding and performs sequence alignment and high-speed clustering based on that geometric representation.