Decombinator

Decombinator assigns gene sequences in T-cell receptor (TcR) repertoires using a finite-state automaton approach to rapidly categorize Illumina short-read high-throughput sequencing data.

Key Features:

• Unique Identifier System: Employs a five-item identifier that uniquely and unambiguously defines each TcR sequence.
• Finite-State Automaton Mapping: Uses finite-state automata to map Illumina short-read TcR sequence data to unique identifiers.
• Error Handling: Includes an extension to accommodate single-base pair mismatches arising from sequencing errors.
• Performance and Accuracy: In tests on in silico and published human TcR-β sequences, demonstrated more than two orders of magnitude faster performance than classical pairwise alignment algorithms and achieved >88% accuracy with up to 1% introduced error rates.

Scientific Applications:

• Antigen-Specific Receptor Analysis: Processes large TcR datasets to enable exploration of antigen-specific receptor breadth and depth.
• V and J Usage Bias: Identifies biases in V and J gene usage within human peripheral blood T-cell repertoires independent of antigen exposure.
• Repertoire Size and Diversity: Characterizes repertoire size and diversity, highlighting challenges in achieving comprehensive descriptions of TcR repertoires.

Methodology:

Applies finite-state automata to map Illumina short-read TcR sequences to five-item identifiers with an extension to accommodate single-base pair mismatches; implemented in Python 2.6.