The Starr Report famously identifies the semen found on a certain blue dress as President Clinton's. An appended FBI lab report provides details of the DNA analysis, meaning that President Clinton's so-called genetic fingerprint is public information (available, in fact, to anyone who clicks here). This week the Drudge Report and New York Post reported that the Star tabloid is using this information to test the paternity of 13-year old Danny Williams. Danny's mother, a former Arkansas prostitute, has long claimed that Clinton is the boy's father. (Slate, of course, does not traffic in such slime. We do, however, report on what's in publications without our high standards. See Slate's "Keeping Tabs" and "International Papers.") Can Clinton's paternity or lack thereof really be established from the information in the Starr report, or do you need a sample of the semen itself?
The human genome--which contains all the information for building one's body and brain--can be thought of as a sentence several billion letters long. Scientists will someday be able to infer an enormous amount of personal information about health and psyche from this sentence, which is unique to every person. But the so-called "genetic fingerprint" includes a trivial fraction of that information. It's enough to distinguish one person from another, but not enough to clone Bill Clinton from the Starr report.
Specifically, the FBI's "genetic fingerprint" consists of what the sentence says in seven places (each around 600 letters long). If two samples of DNA are different in any one of the seven places, we can be positive that they came from different people. If the samples are identical in these seven places, they are assumed to have come from the same person. It's possible that samples identical in the seven places are different elsewhere and therefore belong to different people. But if the test is done right, the chances are one in several billion. (That is still enough for some defendants, like OJ Simpson, to claim that a DNA match isn't conclusive.) Paternity tests are a little more complicated than ordinary DNA matches because only half of a child's DNA comes from the father. But a lab that has the mother's, child's, and putative father's DNA can piece together the puzzle, and these tests are remarkably accurate as well.
The FBI performed two genetic fingerprinting tests on the president's DNA. The Starr report, for unexplained reasons, gives data only for the less specific of the two tests. In fact, this test is imprecise enough that it would probably not be persuasive to a judge. Dr. George Riley of Genelex--a well-regarded forensic DNA lab--calculates that the genetic fingerprint given in the Starr report will most likely yield a so-called "paternity index" of only 20 to 30. In other words, a positive test would mean that President Clinton is only 20 to 30 times more likely than a random Caucasian male to be Danny Williams' father. This would be suggestive but not conclusive. (50,000 white men in Arkansas would get the same test results.) The legal threshold is 100. At best the genetic fingerprint contained in the Starr report can only yield a paternity index of 120. The actual value depends on the mother's DNA.
Is everybody's DNA an open book like this? The answer is yes, more or less. Even if Clinton's genetic fingerprint weren't published, if Bobbie Ann Williams filed a paternity suit, a judge would probably compel Clinton to give her lawyers a blood sample. The law does not give individuals much privacy protection over DNA information. For instance, if you lick a stamp and mail it, the letter's recipient is permitted to perform a genetic analysis on saliva stuck to the stamp. The same is true of skin cells left on a drinking glass or hair left in a shower drain. In fact, since humans are shedding cells all the time, it is unlikely that an ordinary person could ever keep his or her DNA to his or herself.
Explainer thanks Harvey Silverglate of Silverglate & Good of Boston, Mass.