TCRpair

TCRpair predicts functional pairings between HLA-A*02:01-restricted T cell receptor (TCR) α and β chains by analyzing CDR3 and flanking amino acid sequences without requiring the cognate peptide sequence.

Key Features:

• Predictive Capability: Predicts whether α and β chain pairs form functional TCRs specific to HLA-A*02:01 using hypervariable CDR3 sequences.
• Consideration of Flanking Amino Acids: Incorporates amino acids flanking the CDR3 regions to improve predictive accuracy, reporting an area under the curve (AUC) of 0.71.
• Implementation: Implemented in Python using TensorFlow 2.0.

Scientific Applications:

• High-throughput Characterization: Supports high-throughput screening of therapeutic TCRs for development of cancer immunotherapies and other immune-based treatments.
• Functional Pair Identification: Enables identification of functional α/β chain pairings to support research into T cell biology and antigen recognition.

Methodology:

Leverages machine learning to analyze TCR α and β chain sequences with focus on CDR3 regions and adjacent flanking amino acids; implemented in TensorFlow 2.0.