nf-core airrflow

nf-core airrflow processes Adaptive Immune Receptor Repertoire sequencing (AIRR-seq) data to analyze B-cell and T-cell receptor repertoires for studies of adaptive immune responses in health and disease.

Key Features:

• Comprehensive Data Handling: Processes targeted amplicon bulk V(D)J sequencing (multiplex PCR or 5' RACE), single-cell V(D)J sequencing from 10xGenomics libraries, and assembled reads from bulk and single-cell datasets.
• Integration of Immcantation Framework: Uses the Immcantation toolset for processing from raw read handling through clonal inference adhering to BCR and TCR analysis best practices.
• Scalability and Portability: Implements workflow management with Nextflow to enable parallelization, scalability, and execution across diverse high-performance computing environments.
• Reproducibility through Containerization: Supports Docker and Singularity containers to provide reproducible software environments.
• Validation and Application in Research: Validated on simulated sequencing data with errors and on real datasets, including analysis of AIRR-seq data from 97 COVID-19 infected individuals and 99 healthy controls to assess convergent antibody responses to SARS-CoV-2.

Scientific Applications:

• Immune repertoire profiling: Enables analysis of B-cell and T-cell receptor repertoires to study immune system state in infectious disease, autoimmunity, and cancer.
• High-throughput cohort analyses: Supports large-scale AIRR-seq studies, including cohort-level comparisons such as SARS-CoV-2 infected versus healthy control datasets.
• Reanalysis and validation: Facilitates reanalysis of publicly available AIRR datasets for validation of findings and discovery of repertoire-level patterns.

Methodology:

Nextflow orchestrates the workflow; the Immcantation toolset performs raw read processing through clonal inference; supports bulk and single-cell V(D)J reads (multiplex PCR, 5' RACE, 10xGenomics) and assembled reads; uses Docker/Singularity containers and has been validated on simulated sequencing data with errors and real datasets.