Genomic Variation in Africa: Insights from the DDD-Africa Study

When interpreting genomic variation in diverse populations one is almost guaranteed that uncertainty will be a factor to deal with at times. The limitations to knowledge and resources representing patients from understudied population groups influence our ability to interpret genomic variation and is one of the main reasons why initiatives to improve datasets are so important.

Exome sequencing is a proven approach to increase clinical utility and diagnostic yield for developmental disorders, but there is a knowledge gap related to how best to tailor genomic screening to multi-ethnic populations. This is the focus of the Deciphering Developmental Disorders in Africa (DDD-Africa) research study, a collaborative project between South Africa and the Democratic Republic of Congo. The rationale of DDD-Africa is to map the genetic aetiology of DD in Africans, whilst exploring the considerations and challenges to implementation of genomic medicine efforts in diverse populations. The following lessons learnt from DDD-Africa showcase real-life examples of how researchers deal with uncertainty in practice, and how diversifying genomic resources can have a positive impact on variant interpretation.

Lesson 1 – A diverse patient’s clinical features might not match typical European-based dysmorphology

A 9-year-old female participant of North African descent with Moderate ID, a ventricular septal defect, hypotonia, dysmorphic features, macrocephaly, and short stature was enrolled in DDD-Africa. Clinical features were thought to be indicative of Cornelia de Lange syndrome, but exome analysis revealed an ARID1B frameshift deletion (ARID1B ENST00000346085.10:c.2856​del (p.Pro953GlnfsTer44)) and the variant was classified as likely pathogenic. Mutations in ARID1B are known to cause Coffin-Siris syndrome (OMIM entry #135900) a rare, clinically heterogeneous disorder. This patient’s clinical features could be utilised to update and refine dysmorphology analysis for African patients.

A further example is of three patients who were all found to have pathogenic variants in the CHD7 gene. Mutations in CHD7 are known to cause CHARGE syndrome (OMIM entry #214800). Only one of these patients displayed classical clinical features of CHARGE syndrome (CHD7 NM_017780.4:c.232C​>T(p.Gln78Ter)). Information from diverse patients can help to expand our knowledge of the phenotype of rare disorders, and to revise diagnostic criteria, illustrating the benefit of diversifying the genomic knowledgebase to include under-represented population groups.

Lesson 2 – Variants and disorders known to be very rare in European populations, may be more common in diverse populations

PGAP3-Congenital Disorder of Glycosylation (PGAP3-CDG, OMIM entry #615716) is a rare genetic disease with features of developmental delay, hypotonia and ataxia, as well as an increased level of the enzyme alkaline phosphatase (AP) in the blood (hyperphosphatasia). Given how rare this disorder is (only 30 cases have been published in the medical literature by 2021) it was surprising to find five patients in the DDD-Africa cohort with the same variant in the PGAP3 gene, which was classified as likely pathogenic (NM_033419.5: c.557G>C (p.Arg186Thr)). Along with other recent cases from other African cohorts (Bezuidenhout et al., 2020), the recurrence of this variant in unrelated families, in the absence of consanguinity, suggests the possibility of a founder effect in this diverse population group and brings about new knowledge for clinicians and scientist working in the field to consider.

Lesson 3 – Expect many variants of uncertain significance due to the gap in the knowledgebase

As has been demonstrated in many other studies of diverse populations, the DDD-Africa study found a much higher rate of variants of uncertain significance (VUS) than what is observed in European cohorts. This influences the utility of genetic testing in diverse populations, a caveat that will only be mitigated by initiatives aimed at improving the diversity of publicly available datasets that can be used for variant interpretation.

Variant interpretation in unique and diverse populations can lead to many new discoveries, but also new uncertainties, as shown here. Overall, these investigations in diverse populations will broaden our understanding of disease and ultimately lead to genomic medicine applications for all global populations.