New tool for African genomics

Genome-wide association studies (GWAS) in Africa are not as straightforward as such research among populations of people of European or Asian ancestry, because African populations are much more genetically diverse. But researchers studying malaria resistance in western Africa say they've found a way to make African GWAS work. An international team of researchers report in this week's issue of __Nature Genetics__ that it's possible to perform genome-wide studies to probe for genes behind disease in the highly diverse African genomic landscape by combining GWA data, which identify disease-associated regions of the genome, with more specific, robust sequencing data -- like those gathered by the 1000 Genomes Project. "We need to investigate the African population, but it is intrinsically more difficult to study because of its genetic heterogeneity," said linkurl:Pablo Gejman,;http://nuin.northwestern.edu/nuin-content/faculty-records/pablo-v-gejman/?searchterm=gejman director of the Center for Psychiatric Genetics at the North Shore University Health System in Evanston, Ill. The study "shows a sensitive way to overcome the limitations," Gejman, who was not involved with the study, added. GWA studies are typically easier to interpret in people of European or Asian descent, because in those populations, it is more likely that a particular part of the genome will correlate in a predictable way to an area of interest -- such as a gene linked to a disease. As a result, researchers can make better inferences about the genetics behind disease in these populations while genotyping a relatively small fraction of all the variable positions in the genome. In more-diverse African populations, genetic correlations are generally weaker, and this "weak linkage disequilibrium" makes it more difficult to use GWA data to pinpoint which region of the genome correlates to a particular disease. The complex computational method used to combine genotyping and sequence data "helps us to see a way forward to deal with these challenges," linkurl:Dominic Kwiatkowski,;http://www.sanger.ac.uk/Teams/faculty/kwiatkowski/ a human geneticist at the Wellcome Trust's Sanger Institute and lead author of the paper, told __The Scientist__. Kwiatkowski and his coauthors genotyped thousands of Gambian children, probing for novel malaria-resistant genes that may help researchers develop new malaria vaccines. The children in the study were all from a 64 square mile kilometers region of The Gambia known as Kombos. Four separate ethnic groups -- Mandinka, Jola, Wolof and Fula -- accounted for the majority of the study subjects. The team located a specific hemoglobin gene variant that likely plays a role in conferring some resistance to deadly forms of malaria. According to Kwiatkowski, his team's ability to locate that gene variant was vastly improved by sequencing that specific locus in just 62 Gambian people, and combining that sequence data with the broader GWA data. "When we combine the sequence information with the genotyping information we overcome the effects of the weak correlation and the signals we get line up very specifically with the [hemoglobin gene variant]," Kwiatkowski said. As an added bonus, the combinatorial computational method used by the team might enable GWA studies to pinpoint causal genetic variants (such as disease-linked genes) in Africans even more readily than standard GWAS can in Europeans or Asians. European and Asian genomes display a higher degree of correlation between different regions -- something that increases the predictive power of GWA studies in these populations, but also makes identifying the specific location of a causal variant somewhat tricky. Not so in Africa: Kwiatkowski and his colleagues were able to pinpoint the precise genomic location of the malaria resistance hemoglobin variant in study subjects. "In Africa it's much better because now the causal variant shines out brightly at a specific point," he said. Gejman, who conducts genome-wide association studies on schizophrenia in African American populations, noted that the method presented by Kwiatkowski and his team could also help with his work. "The African American population is an incredibly heterogeneous population to study as well," he said. "This study underscores potential difficulty in interpreting GWAS data in African American samples." **__Related stories:__***linkurl:Africa needs basic science;http://www.the-scientist.com/article/display/54777/ [July 2008]*linkurl:We can eradicate malaria: report;http://www.the-scientist.com/blog/display/54422/ [12th March 2008]*linkurl:The genomics of ethnicity;http://www.the-scientist.com/blog/display/55520/ [24th March 2009]

New tool for African genomics

Interested in reading more?

Become a Member of