California's growing DNA library raises questions of its limits

SACRAMENTO — When Donald Carter was booked into the Sacramento County Main Jail in March for allegedly possessing cocaine, he submitted a saliva sample from inside his cheek along with his fingerprints. His charges were dismissed April 10, and he went to a court-ordered drug treatment program.

In accordance with a newly implemented law, Carter's saliva sample was sent to the Department of Justice forensics lab in Richmond. Since January, DNA swabs from everyone arrested on suspicion of a felony in California – not just those convicted of such crimes – have been sent there.

DNA in Carter's swab matched DNA evidence collected 20 years ago at a Sacramento home where 80-year-old Sophia McAllister was murdered. On June 7, Carter was charged with her killing and special circumstances including burglary, robbery and rape.

Since voters in 2004 approved Proposition 69, allowing authorities to collect DNA from people who have been arrested for crimes, the state's library of DNA samples – and the potential to match suspects to specific crimes – has grown exponentially.

But as cases involving DNA matches increase, so does scrutiny of what defines a match.

"If it's a nice clean profile, then the numbers are pretty incredible," indicating it's a probable match, said Ruth Ballard, DNA/biology adviser for the UC Davis Forensic Science Graduate Program and a professor in biology at California State University, Sacramento.

But many experts agree that a partial DNA profile is much harder to evaluate.

"It's those kind of samples that really end up being argued in court," said Ballard, who has served as an expert witness for prosecutors and defense attorneys on DNA evidence. "It's a rational disagreement; it's an area where experts can't decide yet."

DNA holds billions of genetic markers that determine every aspect of a person, from hair color to whether they think brussels sprouts taste good or bad. Criminologists look at only 15 markers to show the molecular difference from one person to the next. They then create a statistic that shows how likely it is that someone else might have the same DNA profile.

"If you have a full profile, and you're looking at all 15 of these markers, when you run the numbers on that, they are so incredibly low, the probability of another person having the same markers as you do is so unlikely, the crime labs are reporting it as a match," Ballard said.

Evidence from cold case homicides, especially those like McAllister's from decades ago, may be degraded. Sometimes only a partial DNA profile can be pieced together, which means that fewer than those 15 points are found.

In these cases, a partial match will pop up in the database, and the standards for how many points must match vary. In California, someone might match with only seven of the points flagged in the database. The FBI flags DNA when at least 10 markers match up.

In other cases, a sample from the victim may contain mixed DNA, and it can be a challenge to determine which DNA belongs to whom.

read the full story at

Related stories from McClatchy DC