Detailed Analysis of Japanese Population Substructure with a Focus on the Southwest Islands of Japan

Takeshi Nishiyama,Haruo Mikami,Ryosuke Ando,Hideshi Niimura,Hirohisa Kishino,Hirokazu Uemura,Kenji Wakai,Keizo Ohnaka,Naoyuki Takashima,Nagato Kuriyama,Sadao Suzuki,Nobuyuki Hamajima,Yin Guang,Hideo Tanaka,Mikako Horita,For The J-Micc Study Group ,Keitaro Mastuo

doi:10.1371/journal.pone.0035000

Abstract

Uncovering population structure is important for properly conducting association studies and for examining the demographic history of a population. Here, we examined the Japanese population substructure using data from the Japan Multi-Institutional Collaborative Cohort (J-MICC), which covers all but the northern region of Japan. Using 222 autosomal loci from 4502 subjects, we investigated population substructure by estimating FST among populations, testing population differentiation, and performing principal component analysis (PCA) and correspondence analysis (CA). All analyses revealed a low but significant differentiation between the Amami Islanders and the mainland Japanese population. Furthermore, we examined the genetic differentiation between the mainland population, Amami Islanders and Okinawa Islanders using six loci included in both the Pan-Asian SNP (PASNP) consortium data and the J-MICC data. This analysis revealed that the Amami and Okinawa Islanders were differentiated from the mainland population. In conclusion, we revealed a low but significant level of genetic differentiation between the mainland population and populations in or to the south of the Amami Islands, although genetic variation between both populations might be clinal. Therefore, the possibility of population stratification must be considered when enrolling the islander population of this area, such as in the J-MICC study.

Highlights

Uncovering population structure is a crucial step in properly conducting association studies because neglecting to correct for population structure can lead to both false positive results and failures to detect genuine associations [1,2,3]
Our study has clearly shown that both the Amami and Okinawa Islanders are genetically differentiated from the mainland Japanese population
This problem arises because allele frequencies can differ between the Amami and mainland population, and the two those designed for genome-wide association study (GWAS)

Summary

Introduction

Uncovering population structure is a crucial step in properly conducting association studies because neglecting to correct for population structure can lead to both false positive results and failures to detect genuine associations [1,2,3]. Ainu and Ryukyu peoples, are recognized as remnant populations descended from the first ancestral group [5,6,7,8,9,10] These peoples inhabit both ends of the Japanese archipelago: the Ainu people live on the northern island of Hokkaido, and the Ryuku people live on the southernmost islands, called Japan’s Southwest Islands, including the Okinawa Islands (Figure 1). Genetic differentiation has not been well examined along the Southwest Islands between Okinawa Islands and the Kyushu (Japan’s southernmost mainland). Previous studies [5,6,7,8,9,10] used only the inhabitants of Okinawa Island, the geographical and political center of the Southwest Islands (Figure 1), as a representative sample of the Ryukyu population. The proponent of the dual structure hypothesis suggests that the Amami Islanders should be included in the Ryukyu population (Figure 1) [5]; a detailed analysis of the Amami Islanders has not yet been conducted

Objectives

Methods

Results

Conclusion