Family members' DNA may lead investigators to the answers, but using it as a forensic technique brings up some troubling questions.
In the end, it was a slice of pizza that sealed his fate: In a scene straight from CSI, a police officer disguised as a waiter retrieved the partly eaten crust and tableware abandoned by the suspect on a restaurant table. Soon, tests confirmed that the DNA in saliva on the pizza matched evidence from a crime scene. That suspect, Lonnie Franklin Jr., is accused of being the serial killer nicknamed the “Grim Sleeper,” who left a trail of at least 10 victims in Los Angeles over a 25-year span. He was arrested in July and has pleaded not guilty to murder charges.
Italian food may have been Franklin’s downfall, but a less common, and more controversial, forensic technique led police to the suspect in the first place. Known as familial DNA searching, it scours existing DNA databases for partial matches that suggest an unidentified suspect may be a close relative of someone in the database. In the case of the Grim Sleeper, such a search pinpointed DNA that partially matched DNA found at one of the killer’s crime scenes, though it belonged to a man too young to have committed some of the murders. Further investigation revealed a probable suspect: that young man’s father, Lonnie Franklin Jr. His capture marks the first time that the technique – currently employed only by California and Colorado – has been used in the United States to help solve a homicide case.
“For those of us who support the judicious use of familial DNA searching in the US, this case is the Holy Grail we’ve been searching for,” says Frederick Bieber, a medical geneticist at Brigham and Women’s Hospital and associate professor of pathology at Harvard Medical School. Bieber and other proponents of familial searching hope the apparent success of the Grim Sleeper case will motivate more states, including Massachusetts, to create a policy that gives investigators permission to implement the technique. But whether Massachusetts will eventually do so is uncertain. The Massachusetts State Police crime lab has a new director, Guy Vallaro, a forensic toxicologist, and he says there are no immediate plans to implement the type of DNA database searching that California and Colorado are using. “California has done a very thorough job of ensuring that they have all the quality measures in place for familial searching,” he says. “But that takes a lot of time and money.”
Familial DNA searching and a related technique called case-specific partial matching draw on basic genetic truths: We receive half of our DNA from our mother and half from our father. No two people, save identical twins, have the exact same DNA sequence, but first-degree relatives like parents, children, and siblings share similar genetic patterns. When compared in the lab, DNA samples belonging to a father and son, for example, aren’t an exact match, but the similarities signal a close familial relationship.
It’s this concept of kinship analysis that informs paternity tests and can help identify victims of crimes and natural disasters. It was, in fact, Bieber’s work with New York investigators after the terrorist attacks of September 11, 2001, that accelerated his thinking on the potential uses for kinship analysis. Now one of this country’s most vocal champions of familial DNA searching, Bieber was part of a panel of scientists who consulted on how DNA samples from 9/11 victims’ families were used to help identify remains when no DNA sample from the missing person (from a toothbrush, for example) was available. “It made sense,” he says, “that we should make more effective use of kinship analysis methods in criminal investigations too.”
A crude version of familial DNA searching was already being used in the United Kingdom to help solve violent crimes, with some success. In one case, the technique had helped solve the murder of a 20-year-old Welsh woman, when, in 2003, 15 years after the crime, a search of the UK National DNA Database found a partial match between DNA collected at the crime scene and the genetic profile of a teenage boy who had provided a DNA sample when he was convicted of an unrelated offense. As with the Grim Sleeper’s son, the boy was too young to have committed the crime, but the information led investigators to the killer, his paternal uncle. Three years later, the same technique helped nab James Lloyd, a prolific rapist in South Yorkshire, England, who had eluded detection for two decades until a familial search brought police to his door.
In 2006, Bieber and David Lazer, then an associate professor of public policy at the Kennedy School (he’s now at Northeastern University), and Charles Brenner, then a visiting scholar at the public health school at the University of California, Berkeley, described these cases in a report published in the journal Science. What caught the attention of scientists, law enforcement, and the media were the statistics. Using a computer model, the researchers estimated that, by putting familial DNA searching to use in the United States, investigators could increase their leads in criminal cases by 40 percent by searching within a local, state, or national database.
That number has impressed some in law enforcement, among them Rockne Harmon. A retired Alameda County prosecutor and current legal consultant, he was a driving force behind the use of familial DNA searching in California, lobbying state Attorney General Jerry Brown and others for its approval in 2008. “This is just another tool to help us exploit DNA evidence to its fullest potential,” he says. “Why shouldn’t we take advantage of that?”
Yet critics of familial searching say the technique raises a number of troubling issues. Forty-seven states require that people provide a DNA sample after being convicted of a felony, some require it upon conviction for a misdemeanor, and a growing number of states collect samples from anyone who has even been arrested in connection with a felony crime. The state databases where those samples are compiled are heavily weighted with minorities, notes UCLA law professor Jennifer Mnookin, because those minorities have higher rates of arrests and convictions. (The reasons for those rates are, of course, much debated.) “If we search databases that are already filled with more African-Americans and Latinos than whites, we’re going to be more likely to find racial minorities that commit crimes,” she explains. “And because the populations contained in the databases aren’t race- or class-neutral, it’s not a matter of luck if the relatives of those populations are affected.”
While he shares those concerns, Bieber points out that in individual cases, DNA itself “is race blind. We don’t know or care whether the owner of that DNA is black, white, or green. Labs are searching every profile in the database regardless of their race or ethnicity.” The odds such a search will lead investigators to the relative of a known offender are high: A Department of Justice survey suggests that 46 percent of jail inmates have another close family member who’s also been incarcerated.
Others argue that DNA databases are a reflection, not a cause, of racial inequalities. “If this information is going to be useful for solving a crime, should we not use the data because they’re demographically unrepresentative?” asks Lazer. “The real issue that we should be wringing our hands about as a society is why African-Americans have conviction rates around seven times that of whites to begin with.”
Another concern about familial DNA searching is that it puts criminals’ relatives under genetic surveillance – possibly tracked for life and subjected to police harassment – simply because their close family member, whose DNA is on record, may have committed a crime. “This approach can cast a net of suspicion over entire families, based on nothing more than partial DNA matches,” says Larry Tipton, head of the Dedham office for the Massachusetts Committee for Public Counsel Services. “It allows genetics to take precedence over everything else.”
The concept “sounds chilling in theory,” admits Harmon. “But in reality, it’s just not true. In California, the cops aren’t even aware of the results of a familial search until scientists have narrowed down the suspects even further.”
Familial searching involves computerized comparisons of DNA samples, then a ranking in order of the likelihood that the known offender profile is genetically related to the unknown profile generated from crime scene evidence. In the UK, the technique stops there, with lab workers providing law enforcement with a list of partial matches. By chance alone, unrelated people can have similar genetic material, so it’s possible that these matches are meaningless false positives. In the case of James Lloyd, police first pursued the partial matches geographically closest to the crime scenes, literally knocking on doors and interviewing people until they stumbled on a viable suspect.
In California and Colorado, however, familial searching involves an additional layer of lab investigation, in which scientists compare Y chromosome markers to further determine whether the DNA from two samples signals a close male genetic relationship. (Although other techniques can tease out close female relatives, they are rarely used, since most violent offenders are male.) Y chromosome testing can remove unrelated profiles, increasing the odds that the DNA samples belong to, say, a father and son, or to brothers. Unlike in the UK, says Bieber, investigators “won’t be knocking on any wrong doors. It almost eliminates the chances of false-positive matches.”
In the case of the Grim Sleeper, familial searching with Y-chromosome analysis suggested a male genetic relationship between the DNA found at the crime scene and the partial match in the DNA database. That information allowed investigators to follow up with old-fashioned police work: surveillance of the suspect and DNA tests of the pizza crust. The technique, say experts, is likely to hold up in court. Last month, the US Ninth Circuit Court of Appeals ruled against a defendant who claimed that providing a sample to a DNA database where it would be available for possible future familial searching could violate both his rights and those of his relatives. The court concluded that familial DNA searching may be no more intrusive than a witness identifying a perpetrator from a photo lineup of similar-looking suspects.
Since Bieber and Lazer’s study in 2006, just California and Colorado have approved the use of familial DNA searching as an investigative tool, although other states have expressed interest. One of the roadblocks to its use, says Bieber, is a general lack of understanding about the technique and lack of communication between scientific and law enforcement staffs, gaps he and Lazer have worked to bridge through educational lectures and conferences.
Questions about the legality of familial searching has also slowed its acceptance. “Lab workers and law enforcement keep saying, ‘We need new legislation to do this,’ ” says Harmon. But “we didn’t change our law in California to allow familial searching. It should be as simple as creating a policy.”
In 2007, a critical report on the Massachusetts State Police crime lab by the consulting firm Vance recommended that the lab make development of a policy for familial searching a priority. Implementation of the technique would only require submitting a policy to Governor Deval Patrick and getting his approval, but change remains slow. Additional resources, such as special software and scientists trained in this type of DNA analysis, would be necessary, according to Vallaro.
Although the lab has no plans to implement familial searching of its criminal DNA databases, it has begun discussions – what Vallaro calls “the first step” – toward instituting a related approach: case-specific partial matching. Case-specific partial matching relies on the similarities between the DNA of close family members but doesn’t involve a search of offender databases. Imagine a case in which a woman is murdered, and blood is left at the crime scene. Investigators ask for voluntary DNA samples from everyone in the household, including the husband, or obtain samples through a search warrant, to look for a complete match or to exclude a person altogether. “A lab worker can examine that DNA and say, ‘Nope, the perpetrator isn’t the husband,’ ” Lazer says. But if the evidence is a partial match to the husband’s DNA, chances are good that it comes from a close male relative – the husband’s brother, for instance.
There’s currently no written policy in Massachusetts to prevent scientists from passing such a clue on to police – but also no policy that allows it. As a result, the practice among State Police crime-lab employees has been not to share partial matches with investigators. Vallaro says he has asked the state’s Forensic Science Advisory Board to discuss a possible policy that would allow case-specific partial matching.
Until then, proponents of familial DNA searching and partial matching worry that concerns about public policy and civil liberties may delay implementation of a tool that has the potential not only to catch criminals but also to prevent more victims of violent crime. Meanwhile, Bieber and others say the Grim Sleeper case has set an example of how the technique should be used. “California has done things right from the beginning,” says Bieber. “Law enforcement reserved familial searching for violent crimes where other investigative methods had failed, and lab employees made sure that the DNA profile data remained anonymous as long as possible.”
Mnookin agrees: “I’ve generally been pretty impressed by California’s policy. I hope that when other jurisdictions follow suit, they’ll remember these lessons.”
Jessica Cerretani is a freelance writer in Dorchester. Send comments to email@example.com.