Murdoch University Research Repository

Welcome to the Murdoch University Research Repository

The Murdoch University Research Repository is an open access digital collection of research
created by Murdoch University staff, researchers and postgraduate students.

Learn more

Bayesian population genetics and the human history of China

Black, M.L., Wise, C.A., Wang, W. and Bittles, A.H. (2003) Bayesian population genetics and the human history of China. In: 11th Conference of the International Society for Computational Biology, June 29 - July 3 2003, Brisbane, QLD.


Throughout recorded Chinese history, regions of the country populated by persons of non-Han ancestry fluctuated significantly in numbers and in terms of their political and commercial influence. Many of these communities had moved from their homelands, settled in China, intermarried with Han Chinese and subsequently formed separate minority populations. Thus the human history of China in large part reflects the impact of migration, population admixture and community endogamy. To assess the possible consequence of these three factors on the current genetic structure of PR China, DNA samples from the majority Han, the Muslim Hui, Bo'an, Dongxiang and Salar, and the more ancient Kuchong, Miao, and Yao minority populations were studied using uniparental and biparental markers. Bayesian algorithms, such as those employed in the program STRUCTURE, were utilised in processing the generated data. The results were compared to analyses utilising more traditional methods such as AMOVA and F-statistics. With both 'traditional' and Bayesian methods, the uniparental and biparental markers exhibited evidence of female gene flow, community endogamy and differing male ancestral origins in the study populations. The results indicate that single marker systems are of limited value in defining population origins and relationships on a historical scale. It is suggested that the future application and development of Bayesian analysis to the study of human genetic diversity should include modelling that accounts for multiple-system data, and non-genetic factors such as recent historical, demographic and cultural effects.

Item Type: Conference Item
Murdoch Affiliation(s): Centre for Comparative Genomics
Conference Website:
Item Control Page Item Control Page