Mining The Genome
Customers can do several things with the data, once available. For example, they can use it to get information about their geographic ancestry, although this is currently a rough estimate at best, broken down into mongoloid, Caucasian and negroid. However, the precision of this data will improve quickly, once more data is available, according to Hsu. He points to a recent study in which Smithkline Glaxo was able to pinpoint the country of origin for European samples with a fairly high degree of precision.
These tests can yield information about paternity and maternity, data that may lead to some awkward realizations. A child has half of his father's SNPs, and half of his mother's. If the child has SNPs that neither the father or mother has, he is either adopted or the father isn't really the father after all. In addition, a male child's Y chromosome is passed down unchanged from the father, and all children get their Mitochondrial DNA (a specialized piece of genetic information in the portion of the cell that produced most of the energy) directly from their mother. So a son or daughter who innocently gets an extended family tested may find skeletons in the closet.
But the use of SNP data that is most controversial is in its relationship to physical characteristics and predilections. Every day, researchers are discovering new associations between SNPs and characteristics such as physical endurance, as well as to potentially life-changing conditions such as diabetes and cancer. It may not be the SNP itself that causes the problem; but because genes tend to stick together through time, SNPs that are close to other defects can serve as markers for the faulty gene itself.
However, the degree to which they may affect an individual may be very weak, explains David Magnus, PhD, Director of the Center for Biomedical Ethics at Stanford. "Most of the traits that we have information about, from say an array that looks at SNPs... in the vast majority of the traits, those are not very predictive. So even if you have a gene for Type 2 diabetes, or that is positively correlated with heart disease, or for being tall, it contributes so little casually, that it doesn't really tell you much about the likelihood of the actually phenotype." This is in contrast to tests such as those for the BRAC1 and BRAC2 genes, which have a large probabilistic impact on the chances of a woman developing breast cancer. "Now [with the new tests] we're talking about things that have so little impact that it's just swamped by any number of other causal factors, including environmental ones."
Sign up for Computerworld eNewsletters.