Identification of a large set of rare complete human knockouts. P. Sulem1, H. Helgason1,2, A. Oddson1, H. Stefansson1, SA. Gudjonsson1, F. Zink1, E. Hjartasson1, G. Sigurdsson1, A. Jonasdottir1, A. Sigurdsson1, O. Magnusson1, A. Kong1,2, A. Helgason1,3, U. Thorsteinsdottir1,4, G. Masson1, D. Gudbjartsson1,2, K. Stefansson1,4 1) Dept Statistics, DeCode Genetics/Amgen, Reykjavik, Iceland; 2) School of engineering and natural sciences,University of Iceland.Reykjavik, Iceland; 3) Department of anthropology, University of Iceland, 101 Reykjavik, Iceland; 4) Faculty of medicine,University of Iceland,101 Reykjavik, Iceland.

   Mutations that cause a loss of function (LoF) of a gene are the causes of many Mendelian diseases. We sequenced the whole genomes of 2,636 Icelanders and imputed the sequence variants identified in this set into 101,584 additional chip typed and phased Icelanders. A total of 6,795 autosomal LoF variants, 3,979 SNPs and 2,816 indels, were found in 4,924 genes. Most LoF mutations are rare, with 85% having minor allele frequency (MAF) below 0.5%. Only 41% of the LoF variants detected in Iceland are present in the Exome Sequencing Project or dbSNP. Recessive diseases are often caused by losing the function of both copies of a gene - the gene being completely knocked out. Here we focus on variants with MAF below 2%, which is the MAF of mutations causing cystic fibrosis, the most common recessive disease in Europeans. It is usually not necessary to sequence the individual affected by a recessive disease in order to observe the causative mutation since it will be present in the heterozygous state in unaffected relatives. Of the 104,220 genotyped individuals, 8,041 (7.7%) have at least one gene completely knocked out by LoF variants with MAF under 2%. These individuals are homozygotes or compound heterozygotes for LoF mutations in 1,171 genes (6.1%) and only 9.9% are children of parents who are second cousins or closer. We assessed whether a gene being completely knocked out depends on its tissue of expression and found that genes that are highly expressed in the brain are less often completely knocked out. We found that LoF mutations are less frequently observed in the offspring of heterozygous parents than Mendelian inheritance would predict. In particular, homozygous LoF offspring of two heterozygous parents are substantially fewer than expected. Detailed phenotyping of the individuals with the extreme genotype of having a gene completely knocked out offers a way of understanding the function of the completely knocked out genes.

You may contact the first author (during and after the meeting) at