The $1000 genome: why would you do it?

Abstract

The prospect of whole genome sequencing (WGS) at a reasonable cost and with accuracy sufficient for clinical diagnostics is tantalising to laymen and professionals alike. In the presentation I will offer a perspective of an academic medical geneticist directing a comprehensive genetic health service for a small country. The cost of WGS is falling rapidly and when the cost of determining an individual genome sequence falls to a price of around $1000 this technology will be applied to a large number of individuals for research and clinical diagnostics. It is even proposed that determining a person’s genome sequence may in the future become a first step upon entering a health insurance or a clinical facility similar to routine blood tests. We don’t at this junction comprehend the extent of the human genomic variation. In early clinical applications of WGS we are likely to be faced with new genetic variation including millions of single nucleotide polymorphisms (SNP), hundreds of thousands of insertion/deletion polymorphisms and thousands of structural variants. The functional significance of recurrent variations is expected to be sorted out eventually with association studies involving whole genomes sequences. However, given the accuracy of DNA replication and the number of cell divisions between zygote and gametes it is expected that each individual has in the order of 50-100 de novo mutations. Many of these are private and eventually eliminated in future generations. The reliable interpretation of their significance will require a deeper than current understanding of the structure-function and copy number relationships of DNA sequences. In short, the advice to those considering clinical WGS is: Be careful. You will be faced with thousands and thousands of novel sequence variants of uncertain clinical significance. We most often don’t know how to interpret their significance. Take for instance missense mutations. Thousands will be present in a genomic sequence. One can look at whether it is a known mutation in a database, evaluate the change in chemical class of amino acids, analyse whether the change involves a critical site, or if the sequence is conserved by interspecies comparison. Computer programs are available to assist in this work. One can also do segregation studies of the variation in the family. These methods are often not reliable in predicting pathogenicity as will be illustrated with examples from my own practice. WGS by its nature involves diagnostic testing, predictive testing, susceptibility testing, pharmacogenetic testing, and carrier testing all in one. The pre- and post-test genetic counselling will be time consuming and complex as will the postanalytical phase of assigning significance to the findings and reporting them. Besides the health implications, privacy issues and risk of unfair discrimination need to be addressed. The associated work is likely to cost much more than the $1000 for determining the sequence! Clinical WGS will presumably first be applied to individuals with a serious condition, which is presumably genetic, but has resisted conventional diagnostic measures. The need to understand the condition, establish prognosis, contemplate therapy as well as to determine recurrence risks is very strong. Often this need will be stronger than any possible disadvantage of WGS. Medical geneticists are currently regularly faced with distressed families where a diagnosis could not be made. They look forward to adding WGS to their armamentarium.