Multi-omics Studies of Childhood Complex Traits in Diverse Populations

National Human Genome Research Institute

Investigators

Linked publications, trials & patents

Abstract

The overall approach of the lab is to apply modern-day human genomics and genetics tools to population-based childhood disease cohorts in order to better understand the pathophysiology of disease. This approach includes the recruitment of cohorts through international, national, and internal collaborators; the application of DNA sequencing technologies, including whole genome-, whole exome-, RNA-, bisulfite-, long-read, and single cell sequencing; and the development of analytical frameworks that span genetic epidemiology, population genetics, and computational biology. Associations derived from this approach are then functionally validated to develop therapeutic hypotheses for the diseases in question. The 12 months covered by this annual report have included varying applications of the above in three main contexts: Childhood Severe Acute Malnutrition (SAM) Our previous studies identified DNA hypomethylation in children with the more-deadly edematous form of SAM (known as ESAM) relative to the non-edematous form (NESAM). Using our previously recruited cohort of 800 DNA samples from Jamaica and Malawi, evenly divided between ESAM and NESAM, we developed a systematic way of assessing correlation between neighboring sites of DNA methylation. Extended studies conducted this reporting year revealed a highly organized structure of correlated DNA methylation that reflects reported patterns of gene regulation. Comparison of correlation patterns between ESAM and NESAM identified disrupted patterns during acute illness. Collaborative metabolomic studies of ESAM identified differences in one-carbon metabolism (1CM) between the ESAM and NESAM. We also previously demonstrated that genetic variation in genes involved in 1CM are significantly associated with ESAM, although the effect sizes of the variants driving this observation differed between the two populations assessed. During this most recent reporting period, we leveraged differences in genetic ancestry between populations to identify the relative contributions of different ancestries to disease risk. We also confirm evidence for selection at 1CM-associated loci by integrating selection assessments from seven populations across Africa with cohort-based selection signatures. We also obtained serum samples from ESAM and NESAM individuals that were used to confirm disruptions in 1CM in SAM, and isolated adequate amounts of cell-free DNA to genotype them in confirmation of our genetic findings. Finally, we further refined our cellular model of starvation-induced fatty liver - a major pathological feature of SAM that is seen much more often in ESAM - by identifying growth media that appropriately mimicked the restricted nutrient environment of liver cells. We also clarified the key timepoints for in vitro fatty-liver accumulation in liver cells following starvation, and were able to experimentally collect DNA for downstream analyses at the tested timepoints. This model will be used to functionally validate genetic, epigenetic, and metabolic findings, as well as test potential interventions. Modifiers of Sickle Cell Disease (SCD) We have been focused on identifying cis-modifiers of fetal-hemoglobin (HbF) production. HbF is known to improve the severity of SCD and previous studies have identified genetic modifiers of HbF production in the beta-globin cluster on chromosome 11, which houses and controls developmental genes involved in the production of beta-hemoglobin. Using targeted long-read sequencing across the beta-globin locus in 40 individuals with SCD from Nigeria, Cameroon, Kenya, and USA, we identified >4,000 putatively novel structural variants (SVs). Given a high suspicion that some of these common SVs were the result of unrepresented DNA sequence in the existing human genome reference, in this reporting period, we re-assessed genomic variation in region by realigning long-reads to the recently published Telomere-to-telomere reference genome. Doing so reduced the number of SVs by an order of magnitude, with most insertions being resolved as part of the reference genome. Paradoxically, this re-assessment identified more novel single-nucleotide variants as a part of the newly represented genome sequence, and identified a novel SVs adjacent to the beta-globin gene that is enriched among individuals with high fetal hemoglobin levels. Work to validate and publish these findings is ongoing. Genomic characterization of African populations Through our labs interactions with the H3Africa Consortium, we have contributed to ongoing efforts to delineate the landscape of African genetic variation. We have collaborated with external researchers to identify African populations that are under-represented in existing genome variation databases. We have developed a plan for whole-genome sequencing of 1400 individuals recruited from these under-represented ethnolinguistic population groups to extend and expand the landscape of human genetic variation. In the most recent reporting period, we engaged with each of the five major African stakeholders to ensure an ethical partnership and collaboration, inclusive of developing detailed protocols for patient recruitment, sample acquisition and shipping, as well as data dissemination and analysis. We have co-authored manuscripts that speak to the complex ethical, moral, and legal issues raised.