Documentation

BED's chromStart/chromEnd are zero-based and half-open ([chromStart, chromEnd)), unlike the 1-based inclusive convention PlasmidFeature and the rest of Domain\Cloning use ; they are converted once here (1-based start = chromStart + 1, 1-based end = chromEnd - the exclusive 0-based upper bound and the inclusive 1-based one are numerically identical). chromStart must be strictly less than chromEnd : a zero-length BED interval has no meaningful 1-based equivalent (naively converting it would produce start > end, which this codebase's convention reserves for an origin-crossing feature - a silent, wrong reinterpretation this reader refuses to make).

BED carries no feature-type column the way GFF3's third column does, so every feature maps to FeatureType::MISC_FEATURE ; the original chromosome/contig name and score, which have no dedicated PlasmidFeature property, are kept in metadata rather than discarded. Only the first six columns (chrom, chromStart, chromEnd, name, score, strand - "BED6") are read : thickStart/ thickEnd/itemRgb and the block/exon columns ("BED12") describe sub-feature structure this reader does not attempt to reconstruct, the same simplification GenbankPlasmidMapper already documents for a GenBank join() location and GffFeatureReader for a spliced Parent/ID relationship. Only the first 3 columns (chrom, chromStart, chromEnd - "BED3") are required ; a missing name falls back to "chrom:chromStart-chromEnd" rather than being rejected, since BED's own spec treats name as optional. Class BedFeatureReader