The Role of Copy Number Variants in the aetiology of developmental disorders in South Africa – a whole exome sequencing study

Abstract

Developmental disorders are rare conditions, causing a mental or physical impairment. Genetic heterogeneity and highly variable clinical manifestations poses great challenges in the diagnosis of developmental disorders. New sequencing technologies have become an important and cost-effective tool in the diagnosis of rare diseases and results can impact clinical practice significantly. Copy number variants (CNVs) play a major role in the pathogenesis of developmental disorders and thus it is important to investigate the presence of CNVs within these patient cohorts. The introduction of exome sequencing (ES) has allowed the detection of CNVs and single nucleotide variants (SNVs) exome wide with a single test. This is a valuable approach to implement especially in a limited resource setting like South Africa. Despite SNVs being well studied, the incorporation of CNV bioinformatics tools into variant calling pipelines has not been implemented routinely as there is no current gold standard for CNV detection from exome data. Very limited clinical CNV studies have been carried out on next generation sequencing (NGS) data in patients of African ancestry. This study thus aimed to identify the most appropriate bioinformatics approach to detect CNVs from exome data. Subsequently, it was implemented in a developmental disorder variant analysis pipeline for ES data generated by the Deciphering Developmental Disorders in Africa (DDD-Africa) study, to establish the role that CNVs play in this African cohort. The DDD-Africa study recruited and performed detailed clinical phenotyping and ES on 500 African patients with an undiagnosed developmental disorder. Four different bioinformatics tools (CANOES, CLAMMS, XHMM and InDelible) have been applied to a set of samples (N=100) with known CNVs which served as a truth set. Functional equivalence evaluation was carried out in order to identify the most effective CNV calling tool or combination of tools to implement on the DDD-Africa exome dataset. Implementing the chosen tools onto the first batch of ES data consisting of 287 participants, yielded a ~7% diagnostic yield. Integrating CNV detection tools into a standard variant analysis pipeline from ES data can improve diagnostic yield while also promoting an improved cost benefit for ES. These results could not only end the diagnostic odyssey, but may lead to better care and management for families with developmental disorders in Africa.