Abstract
Purpose :
Calculating worldwide carrier frequency (CF) and prevalence per population for different autosomal recessive (AR) inherited retinal disease (IRD) mutations and genes based on data deposited in gnomAD.
Methods :
We created an SQL database including information on sequence variants identified in 180 IRD-causing genes extracted from the gnomAD database that contains genetic information on more than 138,000 whole exomes and whole genomes from various populations around the world. Variants were filtered based on allele frequencies, number of alleles and homozygotes in gnomAD, information in mutations databases, and publications in order to identify pathogenic mutations. We additionally calculated prevalence using an algorithm that is based on the Cartesian product that calculates the likelihood of two individuals to carry a mutation in the same gene and have an affected off-spring in each population.
Results :
We identified 4,035 ARIRD pathogenic mutations and calculated the total CF in the different worldwide populations with the highest value of 1/2.1 individuals in the European population. The mutation that shows the highest CF worldwide is ABCA4-p.G1961E (1/29 individuals), followed by ABCA4-p.Gly863Ala with a CF of 1/102. We identified 81 individuals in gnomAD who are homozygous for IRD-causing mutations and therefore are expected to be affected with the corresponding disease. IRD prevalence is expected to be 1/2,024 with the highest prevalence expected in the East Asian population 1/1,582. Interestingly, in most populations we were able to detect specific genes that were common only in that population except for the European population which did not have specific common genes.
Conclusions :
Our analysis shows that CF of ARIRD mutations worldwide is about 38%. In other words, about 2.9 billion individuals are expected to carry at least one mutation that can cause an ARIRD disease. IRD prevalence is expected to be about 0.04% which means that 3.7 milion individuals are expected to be affected by an AR IRD. These calculations can aid in the identification of the genetic cause of IRDs in newly-diagnosed patients in an efficient way. These analyses and computational methods are a stepping stone for similar research in other diseases. Most of our methods rely on statistical analysis of data and thresholds that can be utilized in every NGS analysis disregarding the type of disease.
This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.