Changes in version 1.19.3 (2026-01-25) o checks conventional base probabilities in long-read data o accuracy comparison vs modkit Changes in version 1.19.2 (2026-01-19) o limit/clip reads to targets during BAM loading o filtering by number of context sites o new defaults for filtering Changes in version 1.19.1 (2026-01-02) o API change for thresholding in all 'generate*Report' functions o filtered out reads are dropped and not counted as hypomethylated anymore Changes in version 1.15.3 (2025-03-05) o test coverage for namespaces o new ref Changes in version 1.15.1 (2024-11-16) o can override alignment endness when loading BAM Changes in version 1.13.4 (2024-10-10) o VariantAnnotation moved to Suggests Changes in version 1.13.3 (2024-10-04) o plotPatterns for pretty plotting Changes in version 1.11.9 (2024-03-21) o uses less memory (due to packed SEQ and XM) Changes in version 1.11.8 (2024-03-12) o stricter filtering in lMHL reports Changes in version 1.11.7 (2024-03-05) o optimised reporting from long-read data Changes in version 1.11.6 (2024-02-29) o float and array tags in simulateBam o long-read data input (not optimised yet) Changes in version 1.11.5 (2024-02-12) o RRBS-ready Changes in version 1.11.4 (2024-02-11) o methylation calls for bsmap Changes in version 1.9.8 (2023-09-29) o creates sample BAMs o linearized MHL Changes in version 1.9.4 (2023-07-03) o methylation calls for bwa-meth, etc Changes in version 1.9.2 (2023-06-21) o both paired-end and single-end alignments Changes in version 1.3.6 (2022-02-16) o significant speed-up (1.3.5) o methylation patterns Changes in version 1.3.2 (2021-12-24) o more efficient data handling (XPtr instead of Rcpp::wrap'ping) Changes in version 1.1.9 (2021-09-19) o very fast end memory-efficient BAM loading using HTSlib - for now reads paired-end BAM only o min.baseq to reduce the effect of sequencing errors o very fast Fisher Exact from HTSlib o old code removed Changes in version 1.1.0 (2021-05-21) o released at bioconductor Changes in version 0.99.0 (2021-04-09) o R>=4.0 for submission o removed unused dependencies o correct work of generateVcfReport (although SNV only) o unmatched reads are at the end of generateBed* output now o compiles and works on Apple Silicon (native ARM64 R) o fully documented methods o fully covered with tests and examples o comprehensive vignettes Changes in version 0.4.0 (2021-03-08) o going public o CX report now includes only the context present in more than 50% of the reads o generateVcfReport (capable of dealing with SNVs only for now) o added documentation to some of the methods o added several examples o added sample data for amplicon and capture NGS o added some tests based on sample data o README.md Changes in version 0.3.9 (2021-01-19) o fast C++ CIGAR parser to lay queries in reference space o new method to extract base frequences: generateBaseFreqReport Changes in version 0.3.7 (2021-01-12) o lots of refactoring again o CX report sub now uses boost::container::flat_map (additional 2x speedup) o removed dplyr as a dependence, whole package uses data.table now Changes in version 0.3.5 (2021-01-09) o lots of refactoring o new method: preprocessBam() to save time on loading/preprocessing o new C++ sub for CX report with std::map summary (5-10x speedup) Changes in version 0.3.2 (2021-01-06) o first attempt to stablilize API (generateCytosineReport and generateBedReport) o temporary method for ECDF (generateBedEcdf) o uploaded to GitHub Changes in version 0.3.1 (2020-01-01) o heavy refactoring, many internal methods added o C++ functions for nearly all bottlenecks (pending fast: cigar, summary, genome loading) Changes in version 0.2.1 (2020-12-21) o made this second iteration of epialleleR a usable package