A new paper published in Science Translational Medicine says that the capacity of personal genome sequencing to predict (and direct steps to prevent) common diseases is limited.
The Predictive Capacity of Personal Genome Sequencing reports the examination ofhealth data from thousands of identical (monozygotic) twins, who have identical DNA sequences, and used mathematical modelling to estimate the capacity of genomics to predict the risk of developing 24 common diseases.
They report that, at best, a clinically significant risk estimate would be found for only one disease, and say that genetic testing ‘will not be a dominant determinant of patient care’ nor a substitute for normal medical approaches to disease risk prediction and prevention.
However, other commentators have expressed reservations over both the methods employed by the study and the conclusions. Notably, the paper does not actually examine any genomic data, simply assuming that the correlation in health outcomes for twin pairs mirrors the predictive capacity of genome analysis, and using questionable modeling.
Moreover, previous (more robust) studies that have examined predictive genomics have already shown that power is limited for common diseases, and geneticists don't dispute this fact. Personalised medicine is intended to use genetic testing to refine and improve, as opposed to replace, traditional healthcare.
Comment: What has really annoyed the genetics community is the media attention this work has attracted, and the potentially harmful impression it may give of medical genomics. Prof Leonid Kruglyak of Princeton University commented: “I don’t see the harm in telling the public yet again that there is no such thing as genetic determinism…But I worry about the message being distorted to mean that genes have no value, or that genetic research is not worthwhile”.