Abstract
BackgroundThe admixed South African Coloured population is ideally suited to the discovery of tuberculosis susceptibility genetic variants and their probable ethnic origins, but previous attempts at finding such variants using genome-wide admixture mapping were hampered by the inaccuracy of local ancestry inference. In this study, we infer local ancestry using the novel algorithm implemented in RFMix, with the emphasis on identifying regions of excess San or Bantu ancestry, which we hypothesize may harbour TB susceptibility genes.ResultsUsing simulated data, we demonstrate reasonable accuracy of local ancestry inference by RFMix, with a tendency towards miss-calling San ancestry as Bantu. Regions with either excess San ancestry or excess African (San or Bantu) ancestry are less likely to be affected by this bias, and we therefore proceeded to identify such regions, found in cases but not in controls (642 cases and 91 controls). A number of promising regions were found (overall p-values of 7.19×10-5 for San ancestry and <2.00×10-16 for African ancestry), including chromosomes 15q15 and 17q22, which are close to genomic regions previously implicated in TB. Promising immune-related susceptibility genes such as the GADD45A, OSM and B7-H5 genes are also harboured in the identified regions.ConclusionAdmixture mapping is feasible in the South African Coloured population and a number of novel TB susceptibility genomic regions were uncovered.Electronic supplementary materialThe online version of this article (doi:10.1186/1471-2164-15-1021) contains supplementary material, which is available to authorized users.
Highlights
The admixed South African Coloured population is ideally suited to the discovery of tuberculosis susceptibility genetic variants and their probable ethnic origins, but previous attempts at finding such variants using genome-wide admixture mapping were hampered by the inaccuracy of local ancestry inference
The South African Coloured population (SAC) is a so-called admixed population that derived its origins from the diverse population groups that settled in the early Cape colony, including the indigenous speaking Africans (San), early European settlers, slaves that were imported from Indonesia, India and other parts of Africa, and South African Bantu-speakers who later migrated to the area
As RFMix was not available at that time, our first step was to compare the accuracies of LAMP-LD and RFMix using a simulated data set of 1500 SAC chromosomes
Summary
The admixed South African Coloured population is ideally suited to the discovery of tuberculosis susceptibility genetic variants and their probable ethnic origins, but previous attempts at finding such variants using genome-wide admixture mapping were hampered by the inaccuracy of local ancestry inference. Previous genetic research has shown that the SAC received ancestry contributions from click-speaking Africans (San), Bantu-speaking Africans, Europeans and South and East Asians, which is consistent with the historical records [1,2,3,4,5]. As the group received contributions from diverse source populations that may differ in their genetic susceptibility to TB, the group is ideally suited to the discovery of TB susceptibility genetic variants and their probable ethnic origins. The process of finding such areas is known as admixture mapping, and this technique relies on the accurate inference of what is known as local ancestry per individual across their genome [10]
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
