Challenges in molecular diagnosis of X-linked Intellectual disability

Abstract

Abstract Background Intellectual disability (ID) affects 1–3% of the Western population and is heterogeneous in origin. Mutations in X-linked genes represent 5–10% of ID in males. Fragile X syndrome, due to the silencing of the FMR1 gene, is the most common form of ID, with a prevalence of around 1:5000 males. Females are usually non- or mildly affected carriers, and in a few rare cases, the only gender affected. Array comparative genome hybridization (aCGH) and next-generation sequencing (NGS) have dramatically changed the nature of human genome analysis leading to the identification of new X-linked intellectual disability syndromes and disease-causing genes. Sources of data Original papers, reviews, guidelines and experiences of the diagnostic laboratories. Areas of agreement Family history and clinical examination still are essential to choose the appropriate diagnostic tests, including, a disease-specific genetic test, aCGH or FMR1 molecular analysis. If negative, NGS approaches like well-defined gene panels, whole exome, or even whole genome sequencing, are increasingly being used, improving diagnostics and leading to the identification of novel disease mechanisms. Areas of controversy The main challenge in the era of NGS is filtering and interpretation of the data generated by the analysis of a single individual. In X-linked cases, assessing pathogenicity is particularly challenging, even more when the variant is found to be inherited from a healthy carrier mother or when a heterozygous X-linked mutation is found in an impaired female. Growing points At present, variant interpretation remains a challenging task, especially in X-linked disorders. We review the main difficulties and propose a comprehensive overview that might aid in variant interpretation. Establishing a genetic diagnosis facilitates counseling and allows better delineation of clinical phenotypes. Areas timely for developing research To improve variant interpretation, there is need to refine in silico predictions with specific criteria for each gene, and to develop cost-effective functional tools, which can be easily transferred to diagnostics.