LDblockR: a cross-platform R toolkit for multi-format linkage disequilibrium analysis and integrated regional genomics visualization

Linkage disequilibrium (LD) analysis is central to candidate-gene refinement, haplotype characterization, marker selection, and the interpretation of genome-wide association studies. However, commonly used LD programs often require format conversion, platform-specific binaries, or separate tools for LD calculation, block detection, regional annotation, subgroup comparison, and publication-quality visualization. We developed LDblockR, a pure-R, cross-platform toolkit that unifies regional LD analysis from VCF/VCF.GZ, HapMap, PLINK BED/BIM/FAM or PED/MAP files, and numeric 0/1/2 genotype matrices. LDblockR calculates dosage- or phased-haplotype-based r² and Lewontin’s D′, estimates haplotype frequencies for unphased diploid genotypes using an expectation-maximization procedure, and supports multiple block definitions, tag-SNP selection, LD-neighbor searches, LD decay, haplotype frequencies, population-group comparisons, and batch regional analysis. Its principal contribution is workflow integration and interoperability rather than a new LD estimator: a single R-native object model links heterogeneous genotype inputs, established LD statistics, regional annotation, and publication-ready visualization. In the bundled phased benchmark of 60 diploid samples and 42 variants (861 unique pairs), LDblockR agreed with LDBlockShow 1.41 to the three-decimal precision reported by LDBlockShow, with all r² and D′ values within 0.00051. We additionally analyze the official TASSEL/rTASSEL maize HapMap tutorial dataset, comprising 281 maize inbred lines and 3,093 genome-wide markers. After allele-consistency, missingness, and MAF filtering, 2,526 markers remain, including 428 SNPs on chromosome 1. Chromosome-wide dosage r² analysis covers 91,378 pairs; a highlighted 1-Mb interval contains 16 SNPs and 120 pairs. Independent pairwise calculations match the archived regional reference table to a maximum absolute difference of 5.01 × 10 −13 . This real-data example demonstrates regional LD analysis on crop genotypes with observed missing calls and allele-metadata conflicts. LDblockR therefore provides a transparent bridge between genotype input, regional LD analysis, and post-GWAS interpretation for plant association genetics and breeding.