Background <p>Long-read sequencing has enabled the generation of high-quality human genome assemblies, but many previous assemblies were based on blood-derived DNA and often relied on limited data types from a single sequencing strategy.</p> Objective <p>This study aimed to generate high-quality phased genome assemblies of a Korean individual using multiple independent long-read datasets produced from a single sequencing platform and to evaluate their utility for chromosome-scale assembly and variant detection.</p> Methods <p>Genomic DNA was extracted from a semen sample of a Korean male. Long-read, ultra-long-read, and chromatin conformation capture sequencing data were generated using Oxford Nanopore Technologies. These datasets were integrated to construct phased genome assemblies, followed by correction of noticeable phasing errors and assessment of assembly continuity, chromosomal representation, telomeric repeat recovery, and variant detection performance.</p> Results <p>The final phased assemblies spanned approximately 2.9&#xa0;Gb and represented 23 pairs of chromosomes with an NG50 of 150&#xa0;Mb. Telomeric repeats were detected at 36 and 37 of the 48 chromosomal ends in the two assemblies, indicating high end-to-end completeness.</p> <p>In addition, we successfully identified structural variants, including small variants. These results demonstrate that combining multiple Oxford Nanopore data types can produce highly continuous and informative phased human genome assemblies.</p> Conclusions <p>We generated high-quality phased genome assemblies of a Korean individual using Oxford Nanopore long-read sequencing data derived from semen DNA. This publicly available genome resource will support broader applications of long-read sequencing in human genomics and variant analysis.</p>