Show simple item record

dc.contributor.authorAyodo, George
dc.date.accessioned2011-08-11T13:44:25Z
dc.date.available2011-08-11T13:44:25Z
dc.date.issued2011-08-11
dc.identifier.urihttp://ir-library.ku.ac.ke/handle/123456789/732
dc.descriptionThesis submitted in partial fulfilment of the requirements for degree of doctor of philosophy in genetics of infectious diseases in the school. RA 644 .M3A9en_US
dc.description.abstractMalaria causes death of millions of people in sub-Saharan Africa and about 80% are children and women under 5 years of age. Infection has therefore exerted pressure on human genome and as a consequence clinical manifestations appear variable in endemic and non-endemic populations. Part of the reason for this epidemiological difference is hypothesized that over the last few thousand years, endemic populations have built up genetic resistance to severe malaria infection. To test this hypothesis, the study searched for evidence of natural selection in malaria exposed and unexposed populations by (a) carrying out a large-scale collection in Kenya of severe malaria cases and controls from the Luo ethnic group and also of population controls from the Masai and Kikuyu ethnic groups, (b) carrying out an association study at 10 genetic variants previously associated with malaria resistance, (b) studying frequency differences across populations to determine which of these variants have been subject to selection for malaria resistance in the past few thousand years, and (c) also studying haplotype and linkage disequilibrium patterns around malaria resistance genes to search for evidence of natural selection. In the Luo case-control samples, the previously described associations at CD36-GT (P value < 0.004) and HbAS (P value = 0.015) were replicated. Strikingly, there was unusually high frequency differentiation of the HbAS and CD36-GT variants in the exposed (Luo and Yoruba) vs. relatively unexposed (Kikuyu and Masai) populations compared to a panel of 1,454 randomly chosen markers that were studied in the same samples (P < 0.00036 and 0.00096 respectively). By statistically combining the case-control association and frequency differentiation statistics, the power of the association analysis was increased by several orders of magnitude (HbAS with P value < 0.0000 18 and CD36-GT with P value < 0.00043), which provides a potential tool for researchers to find risk factors for infectious disease in future. Further assessment o€ haplotype blocks flanking HbAS-T, CD36-G and ICAM-T suggested that exposed and un-exposed populations exhibit different haplotype block patterns, supporting the evidence of natural selection. CD 36GT appears to be under selection in both the Luo and Yoruba ethnic group, whereas HbAS is under selection in the Yoruba ethnic only but not in Luo ethnic groups. These results suggest Yoruba and Luo-perhaps because they are on different sides of the African continent-evolved different genetic response to malaria because they had been exposed to the disease for thousand years. This study has not only developed a novel method to identify malaria variants but has also provided an insight on the possibility of exploiting haplotype block patterns to map causal genesen_US
dc.description.sponsorshipKenyatta universityen_US
dc.language.isoenen_US
dc.subjectMalariaen_US
dc.titleAssociating genetic resistance to plasmodium falciparum malaria infection with ethnic groups residents of malaria endemic and non-endemic regions of Kenyaen_US
dc.typeThesisen_US


Files in this item

Thumbnail

This item appears in the following Collection(s)

Show simple item record