Changes in version 0.99.4 - Added \value section to TCR_summary-class documentation to resolve R CMD check warning about empty/missing \value sections Changes in version 0.99.3 - Addressed Bioconductor reviewer feedback - Updated R dependency to >= 4.6.0 - Converted TCR_summary from S3 class to formal S4 class with show method - Added ORCID to Authors@R - Replaced \dontrun{} with \donttest{} in all examples except runSoNNia (kept as \dontrun{} due to upstream soNNia/numpy incompatibility) - Removed redundant tryCatch/stop() pattern in huggingModel(); replaced with on.exit() cleanup - Extracted shared .run_in_basilisk() helper to reduce code repetition across calculate.* functions - Added @importFrom SummarizedExperiment colData and removed explicit :: qualifiers in runOLGA, runClustTCR, and runEmbeddings - Added n == 0L edge-condition guard for seq.int() in proteinEmbeddings() - Expanded vignette introductions with biological context and detailed prose descriptions for each code section - Removed GitHub installation instructions from vignettes - Added additional unit tests for S4 class structure, edge cases, and input validation Changes in version 0.99.2 - Removed the umbrella roxygen block from R/utils.R - Added skip_on_bioc_build() check directly inside skip_if_no_python() in tests/testthat/helper-immLynx.R Changes in version 0.99.1 - Switched example data from Seurat to SingleCellExperiment object - Replaced Seurat with scran/scater in vignettes - Removed Seurat dependency; functions now use SingleCellExperiment exclusively - Renamed return_seurat parameter to return_input in runMetaclonotypist() Changes in version 0.99.0 - Initial Bioconductor submission - Added runClustTCR() for TCR clustering via clusTCR - Added runTCRdist() for pairwise TCR distance calculations - Added runOLGA() and generateOLGA() for generation probability - Added runEmbeddings() for protein language model embeddings - Added runMetaclonotypist() for metaclone discovery - Added runHLAassociation() for HLA-metaclone associations - Added runSoNNia() for selection inference - Added utility functions: extractTCRdata(), validateTCRdata(), convertToTcrdist(), summarizeTCRrepertoire() - Added huggingModel(), tokenizeSequences(), and proteinEmbeddings() for custom embedding workflows