In the 1997 science fiction thriller Gattaca, Ethan Hawke's character\nfights to make a life for himself in a world where pervasive genetic\ntesting and manipulation has left the imperfect by the wayside. At the\ntime, it was easy to dismiss the film as a futuristic flight of fancy,\nnot an issue that people would be facing in the next 10 years. After\nall, the Humane Genome Initiative already had spent years trying to\nsequence a single man's DNA, and had years left to go.\n\nBut, as often happens in technology, the science of genomics followed\na "hockey stick" curve. The initial breakthrough of a single\nsequenced genome required the development of new sequencing\ntechniques, which have drastically reduced the cost and time required.\n\nCurrently, commercial services will deliver an entire sequenced\ngenome in less than a month, and there is a concerted effort to reduce\nthat price to under $1,000. Meanwhile, several companies offer\nconsumers the ability to get vast amounts of their genetic data for as\nlittle as $400. Personal genomics, so recently science fiction, is now\na click away on the Internet. But while this information may offer\nnew insights for individuals into their heritage and physical\ncharacteristics, the possibility of a Gattaca-like future suddenly\ndoesn't seem so far-fetched.\n\nThe two best-known personal genomic services, 23andMe and deCODEme,\nboth operated in essentially the same manner. After visiting their\nweb site and paying for the testing ($1,000 for deCODEme, $400 for\n23andMe), a kit is sent in the mail. The consumer either rubs a stick\nshaped like a large tongue depressor against the inside of their\ncheek, or provides a (somewhat voluminous) sample of their saliva.\nThe kit is returned to the company, which then processes the sample.\n\nAfter a few weeks, the customer's genetic data becomes available on\nthe website. But what he gets is not his complete genetic\nsequence. Instead, he gains access to as much as a million of their\nSingle Nucleotide Polymorphisms (SNPs). Human DNA is, on the whole, identical from person to person. It is in the small differences\nbrought on by random mutations that all the differences we see between\npeople occur. Many of these differences are a single letter chance in\na portion of DNA, a G instead of a C, for example. Fortunately for the\ncompanies, these single letter changes are relative easy to detect,\nand there are now "Snip Chips" that can detect the values of hundreds\nof thousands of individual SNPs from a single sample at the same time.\n\n"All genetic variation is a result of random mutation," explains Andro\nHsu, Science and Policy Liaison for 23andMe. "A lot of SNPs are\nneutral, in that they don't seem to have any effect on phenotype,\nwhich is the physical characteristics of a person. When you have\nenough of these mutations, and they get spread out to children, then\nyou get a natural set of variation across the population. Over 99.5 percent\nof the genome is identical between human beings, and it's that last\n0.5 percent or so that's different. And SNPs make up a good deal of those\ndifferences."Mining The Genome\nCustomers can do several things with the data, once\navailable. For example, they can use it to get information about\ntheir geographic ancestry, although this is currently a rough estimate\nat best, broken down into mongoloid, Caucasian and negroid. However,\nthe precision of this data will improve quickly, once more data is\navailable, according to Hsu. He points to a recent study in which\nSmithkline Glaxo was able to pinpoint the country of origin for\nEuropean samples with a fairly high degree of precision.\n\nThese tests can yield information about paternity and maternity, data\nthat may lead to some awkward realizations. A child has half of\nhis father's SNPs, and half of his mother's. If the child has\nSNPs that neither the father or mother has, he is either adopted or\nthe father isn't really the father after all. In addition, a male\nchild's Y chromosome is passed down unchanged from the father, and all\nchildren get their Mitochondrial DNA (a specialized piece of genetic\ninformation in the portion of the cell that produced most of the\nenergy) directly from their mother. So a son or daughter who\ninnocently gets an extended family tested may find skeletons in the\ncloset.\n\nBut the use of SNP data that is most controversial is in its\nrelationship to physical characteristics and predilections. Every\nday, researchers are discovering new associations between SNPs and\ncharacteristics such as physical endurance, as well as to potentially\nlife-changing conditions such as diabetes and cancer. It may not be\nthe SNP itself that causes the problem; but because genes tend to\nstick together through time, SNPs that are close to other defects can\nserve as markers for the faulty gene itself.\n\nHowever, the degree to which they may affect an individual may be very\nweak, explains David Magnus, PhD, Director of the Center for\nBiomedical Ethics at Stanford. "Most of the traits that we have\ninformation about, from say an array that looks at SNPs... in the vast\nmajority of the traits, those are not very predictive. So even if you\nhave a gene for Type 2 diabetes, or that is positively correlated with\nheart disease, or for being tall, it contributes so little casually,\nthat it doesn't really tell you much about the likelihood of the\nactually phenotype." This is in contrast to tests such as those for\nthe BRAC1 and BRAC2 genes, which have a large probabilistic impact on\nthe chances of a woman developing breast cancer. "Now [with the new\ntests] we're talking about things that have so little impact that it's\njust swamped by any number of other causal factors, including\nenvironmental ones."GINA To The Rescue?While some see personal genomics as ushering in a new era of\npersonalized health care, it also raises the specter of discrimination on the basis of genetic data. Conceivably, health insurance companies\ncould start to consider a positive SNP test for a disease to be a\npre-existing condition, for example. Or an employer might not hire\nsomeone who carried an increase risk of a mental illness, according to\ntheir genetic data. To answer some of these concerns, Congress passed\nGINA, the Genetic Information Nondiscrimination Act, in 2008.\n\nBut as\nHsu points out, it's only a first step. "As it stands, it only applies\nto health insurance coverage and to employer discrimination based on\ngenetic information. It doesn't currently apply to life insurance or\nlong term care insurance, things that people's decisions might be\neffected by because of genetic information." He also points out that\nGINA only deals with the approval or denial of health care coverage,\nleaving insurance companies and health plans free to charge higher\npremiums for those they deem risky.\n\nThe issue of employment discrimination is particularly complex,\naccording to David Magnus. Imagine that a certain subgroup in the\ngeneral population is likely to become ill in the presence of some\npathogen which is otherwise well-tolerated by people. If there was a\ntest for that variation, a company might very well want to screen\npotential employees for it, if the work environment would expose the\nemployee to the pathogen. "But then that raises questions about who\nhas the authority to make those decisions," Magnus says. "Can you\nforce that testing on people against their will? If they decided that\nthey needed the job, that it was much better paying than any other job\nin town, could they decide that it's worth it to get exposed to the\nrisk? These are all the kinds of concerns that people have been very\nworried about."\n\nMagnus also points out that the medical community may not be prepared\nto deal with this flood of new data. "This is a huge problem, because\nthen what are you supposed to do with this data? If you can't\nunderstand it, and a clinician can't understand it, what are we going\nto do with it? I'm very concerned that people are going to believe\nthat they are at increased risk or decreased risk [for a disease],\nwhen they aren't."Brave New World, V2.0\nAlthough discrimination on the basis of genetic information could be a\nserious issue, there is another use for genetic testing that's\nstarting to emerge, with disturbing implications: designer babies.\nPrenatal screening for genetic abnormalities in high-risk parents,\nsuch as those who carry the genes for Tay-Sachs disease, is\ncommonplace these days; although the moral dilemma that a positive\nresult can create is still a difficult issue. But as more and more\ntraits are testable through SNPs and other genetic markers, it is\nbecoming possible to design the child you want.\n\nIn its more benign form, this takes the form of sampling a cell from\nin-vitro fertilized (IVF) embryos and choosing which one to implant\nbased on the results. So, if the parents want a blue-eyed child, only\nblue-eyed embryos would be implanted. Parents can already choose the\ngender of the child to be implanted, sometimes to prevent a\ngender-linked disease from being passed on, and sometimes to create\n"family balance."\n\nThe more sinister scenario is the use of amniocentesis or ultrasound\nstudies of a fetus to abort undesirable children. This is not science\nfiction; it's a growing problem in India and China. Selective\nabortion of girls has led to an increasingly out of balance gender\nratio, with 20 percent more boys than girls being born in China.\n\nAs the price of genome testing falls, and more and more traits are\nidentified, we face the possibility that other traits, such as blond\nhair or green eyes or height could be similarly skewed. Today, very\nfew children are born by IVF, but the allure of creating the "perfect"\nbaby may be hard to resist.\n\nDavid Magnus points out that some IVF couples already practice a\nversion of this, by selecting egg and sperm donors with traits they\ndesire. But he's less worried about skewing the genome than he is\nabout the welfare of the children, and what is called "The right to an\nopen future." "If you allow prenatal testing, if you allow selection\nof these sorts, if you allow newborn testing, you do worry about the\nkind of impact that it could have on the children, and the likelihood\nthat they will be as free to make decisions about themselves as they\notherwise would. From the kid's point of view, imagine if the parent\nsays 'Look, I got this egg from someone who's very musically inclined...\nwe did a genome scan when you were born and you have several of these\ngenes that have a very small casual associations with musical ability,\nso we expect you to become a very gifted musician.' But the kid wants\nto play football."\n\nMagnus doubts it will become a serious problem though, at least in the\nUnited States. "The idea of a Gattaca world, where most people do IVF\nin one form or another to procreate, I think is very unrealistic. As\nlong as there is alcohol and youth and indiscretion, people will\ncontinue to have children the old fashioned way."