CONICET | Buscador de Institutos y Recursos Humanos

Malaria remains a major worldwide public health problemin spite of numerous prevention and control efforts in recentyears with ∼ 429,000 deaths in 2015, mostly in sub-SaharanAfrica. The key role of genetics factors in disease suscep-tibility and progression is admitted, but molecular basis ofsusceptibility/resistance to malaria has not been elucidated.In the last few years, several genome-wide association stud-ies (GWAS) have been published on severe malaria. Thisapproach, based on high density genotyping arrays, besidesreplicating known associations (HBB, ABO, G6PD), revealednew genes (ATP2B4, FREM3/GYP genes cluster).Until now, there was no GWAS of non-severe malaria, mainlybecause of difficulties inherent to field studies for largecohorts. We present the result of a GWAS of mild malariaattacks, based on two cohorts of children closely followed-upduring their first year of life in South Benin, and genotypedwith the HumanOmni5 array. After quality control, 800 chil-dren were available for analysis.Two different phenotypes were defined to assess the suscep-tibility of children to clinical malaria: the total number ofattacks recorded in the whole follow-up (classic phenotypein association studies on mild malaria) and the recurrenceof attacks. Using entomological, environmental, and behav-ioral informations, both phenotypes were adjusted on individ-ual exposure to malaria vector and on major covariates, usingrespectively a negative binomial regression and a mixed-effectCox model. This last model allowed to incorporate preciselythe time-dependent exposure. The GWAS was performed onthe adjusted phenotypes, with a mixed model for accountingpopulation structure.