Back

Fine-scale Population Structure and Demographic History of Han Chinese Inferred from Haplotype Network of 111,000 Genomes

Lan, A.; Kang, K.; Tang, S.; Wu, X.; Wang, L.; Li, T.; Weng, H.; Deng, J.; WeGene Research Team, ; Zheng, Q.; Yao, X.; Chen, G.

2020-07-03 genomics
10.1101/2020.07.03.166413 bioRxiv
Show abstract

ABSTRACTHan Chinese is the most populated ethnic group across the globe with a comprehensive substructure that resembles its cultural diversification. Studies have constructed the genetic polymorphism spectrum of Han Chinese, whereas high-resolution investigations are still missing to unveil its fine-scale substructure and trace the genetic imprints for its demographic history. Here we construct a haplotype network consisted of 111,000 genome-wide genotyped Han Chinese individuals from direct-to-consumer genetic testing and over 1.3 billion identity-by-descent (IBD) links. We observed a clear separation of the northern and southern Han Chinese and captured 5 subclusters and 17 sub-subclusters in haplotype network hierarchical clustering, corresponding to geography (especially mountain ranges), immigration waves, and clans with cultural-linguistic segregation. We inferred differentiated split histories and founder effects for population clans Cantonese, Hakka, and Minnan-Chaoshanese in southern China, and also unveiled more recent demographic events within the past few centuries, such as Zou Xikou and Chuang Guandong. The composition shifts of the native and current residents of four major metropolitans (Beijing, Shanghai, Guangzhou, and Shenzhen) imply a rapidly vanished genetic barrier between subpopulations. Our study yields a fine-scale population structure of Han Chinese and provides profound insights into the nation’s genetic and cultural-linguistic multiformity.Competing Interest StatementThe authors AL, KK, ST, XW, LW, TL, HW, JD, QZ, XY, and GC work for WeGene (Shenzhen Zaozhidao Technology Co. Ltd. or Shenzhen WeGene Clinical Laboratory).View Full Text

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.