KOREF_S1: phased, parental trio-binned Korean reference genome using long reads and Hi-C sequencing methods

Abstract

Background KOREF is the Korean reference genome, which was constructed with various sequencing technologies including long reads, short reads, and optical mapping methods. It is also the first East Asian multiomic reference genome accompanied by extensive clinical information, time-series and multiomic data, and parental sequencing data. However, it was still not a chromosome-scale reference. Here, we updated the previous KOREF assembly to a new chromosome-level haploid assembly of KOREF, KOREF_S1v2.1. Oxford Nanopore Technologies (ONT) PromethION, Pacific Biosciences HiFi-CCS, and Hi-C technology were used to build the most accurate East Asian reference assembled so far. Results We produced 705 Gb ONT reads and 114 Gb Pacific Biosciences HiFi reads, and corrected ONT reads by Pacific Biosciences reads. The corrected ultra-long reads reached higher accuracy of 1.4% base errors than the previous KOREF_S1v1.0, which was mainly built with short reads. KOREF has parental genome information, and we successfully phased it using a trio-binning method, acquiring a near-complete haploid-assembly. The final assembly resulted in total length of 2.9 Gb with an N50 of 150 Mb, and the longest scaffold covered 97.3% of GRCh38’s chromosome 2. In addition, the final assembly showed high base accuracy, with <0.01% base errors. Conclusions KOREF_S1v2.1 is the first chromosome-scale haploid assembly of the Korean reference genome with high contiguity and accuracy. Our study provides useful resources of the Korean reference genome and demonstrates a new strategy of hybrid assembly that combines ONT's PromethION and PacBio's HiFi-CCS.

Human ancestry inference at scale, from genomic data

“…2a), CN1 19 (Fig. 2b) and haploid KOREF 20 (Fig. S6) genome assemblies of three individuals places the prevalent ancestry based on highest composition (EUR:54%, EAS:87%, EAS:87%, respectively), consistent with both highest GAI scores, and the published European 18 , Chinese 19 and Korean 20 ancestry for each (Tables S6-S7).…”

“…2b) and haploid KOREF 20 (Fig. S6) genome assemblies of three individuals places the prevalent ancestry based on highest composition (EUR:54%, EAS:87%, EAS:87%, respectively), consistent with both highest GAI scores, and the published European 18 , Chinese 19 and Korean 20 ancestry for each (Tables S6-S7). Further, ntRoot correctly infers EAS and EUR as the two most prevalent super-population ancestries from a synthetic diploid genome admix of HuRef and CN1 (Fig.…”

Small polymorphisms are a source of ancestral bias in structural variant breakpoint placement

“…Advances in long-read sequencing technology coupled with new phased assembly methods are producing more complete and more accurate variant callsets than was previously possible with shorter reads ( Wenger et al 2019 ; Ebert et al 2021 ; Cheng et al 2022 ; Liao et al 2023 ; Rautiainen et al 2023 ). This has facilitated a greater understanding of variation in the past 10 years, especially structural changes in human populations ( Chaisson et al 2015 ; Seo et al 2016 ; Shi et al 2016 ; Audano et al 2019 ; Chaisson et al 2019 ; Beyter et al 2021 ; Ebert et al 2021 ; Kim et al 2022 ; Liao et al 2023 ) and nonhuman species ( Alonge et al 2020 ; Rhie et al 2021 ; Ferraj et al 2023 ; Li et al 2023 ). These advances continue to rival short-read technology by reducing costs, increasing availability, and improving read quality.…”

Pangenome databases improve host removal and mycobacteria classification from clinical metagenomic data

“…The viral library was downloaded on 15 June 2023. The human genome from which the reads were simulated was the Korean reference KOREF_S1v2.1 (RefSeq accession GCA_020497085.1) [ 39 ], with contigs shorter than 10 kbp removed. The bacterial references were obtained by first downloading the bacteria library through kraken, followed by a subsampling due to the size (166 Gb) of the resulting FASTA file.…”