# Construction and integration of three de novo Japanese human genome assemblies toward a population-specific reference

**Authors:** Jun Takayama, Shu Tadaka, Kenji Yano, Fumiki Katsuoka, Chinatsu Gocho, Takamitsu Funayama, Satoshi Makino, Yasunobu Okamura, Atsuo Kikuchi, Sachiyo Sugimoto, Junko Kawashima, Akihito Otsuki, Mika Sakurai-Yageta, Jun Yasuda, Shigeo Kure, Kengo Kinoshita, Masayuki Yamamoto, Gen Tamiya

PMC · DOI: 10.1038/s41467-020-20146-8 · Nature Communications · 2021-01-11

## TL;DR

This paper creates a Japanese-specific human genome reference to improve genetic analysis accuracy for that population.

## Contribution

The study constructs a population-specific genome reference by integrating three de novo Japanese genome assemblies.

## Key findings

- The JG1 reference genome is more contiguous and accurate with Japanese major alleles.
- Using JG1 reduces candidate variant calls in exome analyses while retaining disease-causing variants.
- Population-specific references improve genome analyses for underrepresented groups.

## Abstract

The complete human genome sequence is used as a reference for next-generation sequencing analyses. However, some ethnic ancestries are under-represented in the reference genome (e.g., GRCh37) due to its bias toward European and African ancestries. Here, we perform de novo assembly of three Japanese male genomes using > 100× Pacific Biosciences long reads and Bionano Genomics optical maps per sample. We integrate the genomes using the major allele for consensus and anchor the scaffolds using genetic and radiation hybrid maps to reconstruct each chromosome. The resulting genome sequence, JG1, is contiguous, accurate, and carries the Japanese major allele at most loci. We adopt JG1 as the reference for confirmatory exome re-analyses of seven rare-disease Japanese families and find that re-analysis using JG1 reduces total candidate variant calls versus GRCh37 while retaining disease-causing variants. These results suggest that integrating multiple genomes from a single population can aid genome analyses of that population.

Human reference genomes are typically constructed from few individuals, and are biased towards European and African genomes. Here, the authors assemble three Japanese genomes to create a population-specific reference genome. They then demonstrate improved variant calling from exome sequencing with this reference genome.

## Full-text entities

- **Genes:** F2RL1 (F2R like trypsin receptor 1) [NCBI Gene 2150] {aka GPR11, PAR2}, GNAO1 (G protein subunit alpha o1) [NCBI Gene 2775] {aka DEE17, EIEE17, G-ALPHA-o, GNAO, HG1G, NEDIM}, CYP2U1 (cytochrome P450 family 2 subfamily U member 1) [NCBI Gene 113612] {aka P450TEC, SPG49, SPG56}, CACNA1A (calcium voltage-gated channel subunit alpha1 A) [NCBI Gene 773] {aka APCA, BI, CACNL1A4, CAV2.1, DEE42, EA2}, SPAST (spastin) [NCBI Gene 6683] {aka ADPSP, FSP2, SPG4}, AMPD2 (adenosine monophosphate deaminase 2) [NCBI Gene 271] {aka AMPD, PCH9, SPG63}, glucoside xylosyltransferase 1 [NCBI Gene 100985954], SRY (sex determining region Y) [NCBI Gene 6736] {aka SRXX1, SRXY1, TDF, TDY}, HPX (hemopexin) [NCBI Gene 3263] {aka HX}, CTNNB1 (catenin beta 1) [NCBI Gene 1499] {aka CTNNB, EVR7, MRD19, NEDSDV, armadillo}, AK1 (adenylate kinase 1) [NCBI Gene 203] {aka ADK, Adk1, CNSHA3, HTL-S-58j}
- **Diseases:** hydatidiform mole (MESH:D006828), STS (MESH:D009371), GRC-type AF (MESH:D053591), depression (MESH:D003866), Diplegia (MESH:D002547), thrombophilia (MESH:D019851), type 2 diabetes (MESH:D003924), SNV   10 (MESH:C557827), genetic diseases (MESH:D030342), CHD (MESH:C562377), SV (MESH:D002303), rare diseases (MESH:D035583), HIGH (MESH:D052456)
- **Species:** Pan paniscus (bonobo, species) [taxon 9597], Homo sapiens (human, species) [taxon 9606]
- **Mutations:** c.1276 C > T, c.736 G > A, c.515 + 1 G > A, c.653 C > T, c.1496 G > A, c.651delC, DELTA, c.1724C > T, c.1683 + 2 T > C
- **Cell lines:** PC3 — Homo sapiens (Human), Prostate carcinoma, Cancer cell line (CVCL_0035)

## Full text

_Full body text omitted from this summary view._ Fetch the complete paper as Markdown: https://tomesphere.com/paper/PMC7801658/full.md

## Figures

4 figures with captions in the complete paper: https://tomesphere.com/paper/PMC7801658/full.md

## References

66 references — full list in the complete paper: https://tomesphere.com/paper/PMC7801658/full.md

---
Source: https://tomesphere.com/paper/PMC7801658