Posted on March 20, 2015 at 10:54 AM
Some researchers are at work generating images of people’s faces by relying on DNA samples alone, in a process known as DNA phenotyping. The process involves linking genetic traits and their typical manifestations in traits such as eye color, hair color, and features associated with ancestry. The research is in its early stages, and DNA does its work only in the context of developmental effects, so the facial images are not exact matches, but they are often within the ballpark.
Right now, the interest in DNA phenotyping is mostly forensic, of interest to police agencies trying to identify people who have left no trace of themselves at a crime scene other than blood, semen, or other bodily substance. To help identify suspects through a visual image, private companies already market DNA phenotyping to various police agencies around the country.
It is unlikely that DNA phenotyping will remain of interest only to the law. I expect that some prospective parents might be interested in taking embryonic or fetal DNA to produce a predictive phenotype of what their child might look like as an adult. Will the child look more like the father or the mother? Will the child inherit the father’s generously sized ears or maybe the mother’s aquiline nose? What will its eye and hair color be? Will there be any asymmetry in the child’s face? And what about other body traits? How tall is the child likely to be?
Visual answers to these questions will all depend on the extent to which researchers can meaningfully link genetics to expected physical traits and in a way that is not disturbed by the “noise” of developmental variation. No matter the degree of precision possible, I expect that some entrepreneurial clinicians will be interested in offering answers to these questions, as far as they can, by way of DNA phenotyping.
The DNA sample needed to run such tests could be obtained in a variety of ways. Clinicians could take a cell from an in vitro embryo and characterize its genetics from that sample, as they do now for preimplantation genetic diagnosis. Alternatively, fetal cells might be isolated from amniotic fluid or a pregnant woman’s blood. Mostly, of course, these kinds of efforts are undertaken for diagnostic reasons, to evaluate embryos’ prospects for diseases and disorders. Armed with that predictive information, prospective parents can then select which embryos to use in trying to have children, or which pregnancies to complete.
DNA phenotyping is not at much of a conceptual or emotional remove from these existing diagnostic practices. Right now, courtesy of ultrasound imagery, spectral fetal images decorate parents’ desks and refrigerators; parents share them via social media networks. People seem interested in the emergence of their children’s traits, even from their very earliest stages. DNA phenotyping is in line with this kind of forecasting. Some people might want to know in a general way what their children will look like even before their children look like anything in particular.
For example, knowing what a child will look like might be of special interest to people relying on donor gametes in order to have a child. It is one thing to have a clinic’s gung-ho description of a sperm donor, but it would be another thing to visualize what a child conceived with that donor’s sperm might look like. Many egg donor programs indicate that they are highly selective in the women they accept as donors, but some prospective parents might still want more than that assurance and to have some idea how the donor’s traits will manifest themselves in the child’s appearance.
If we start to use DNA phenotyping at embryonic stages, it may well be that some parents will pass over one embryo in favor of another because of the expected appearance of the future child and adult. They will do so for reasons important to them (“I want a child who looks like me”) and for reasons they represent as important to the child (“I want my child to have the benefits of good looks”). The merit of these reasons is, of course, open to debate. Is it defensible to prefer one possible child over another because of its expected appearance?
Certain bioethicists have counselled that parents should choose – when they are able to do so – traits in children that help them have the best possible lives. Is there any reason to exclude physical appearance from the traits that count as reasons that one life is better than another? If so, how does one establish what appearances offer the prospect of a best possible life? If parents undertake DNA phenotyping after a pregnancy begins and they don’t like the forecast image of their children, would they be justified in terminating that pregnancy?
Bioethics knows the analogues of these questions in other contexts, but DNA phenotyping is poised to sharpen their edges anew.
Timothy F. Murphy is a professor of philosophy in the biomedical sciences at the University of Illinois College of Medicine and the author most recently of Ethics, Sexual Orientation, and Choices about Children (The MIT Press).