A hung jury could not convict a man of murder in 2010, partly because of difficult DNA evidence. A pair of bloody gloves was almost certain to have the victim’s DNA on the outside, but the inside showed a complex mixture with potentially four contributors and a chance that the accused was a minor part of the mix in the right glove.

Before the second trial, a San Diego Police DNA analyst cut open the gloves, took more swabs and again used the combined probability of inclusion (CPI) methodology to determine whether Florencio Jose Dominguez was the one wearing the gloves at the time of the murder of 15-year-old Moises Lopez. The results showed a better likelihood that Dominguez had worn them.

The accused was convicted after three days of jury deliberation. But what was never mentioned in the proceedings was that the DNA mixture protocols in San Diego had changed just days before the second trial—something the jury never heard in court.

The California appellate court has now ordered the convicted murderer to be re-tried, or released—since the DNA evidence presented at the second trial was already “invalid,” the judge ruled.

“It is not disputable that (the analyst’s) testimony at the second trial was based on analyses that were then invalid under his own agency’s new protocols,” San Diego Superior Court Judge Charles G. Rogers ruled on Oct. 6. “Moreover, at no time during his direct and cross-examination at the second trial did (the analyst) disclose that the changes in policy had occurred.”


Lopez was assaulted and executed in public on Nov. 13, 2008 in Mountain View Park. A mother and daughter testified they watched from their kitchen window as a group of men beat, dragged and then gunned down the victim with five shots, according to court records.

DNA evidence on a beer bottle near the scene of the beating closely matched Dominguez’s profile, the records show.

But the connection to the actual crime were the bloody gloves, jettisoned a few houses down from the murder site. Lopez’s bloody DNA was all over the outside, and the four-person mixture was inside.

At the first trial in 2010, the gloves mixtures were analyzed separately, using a technique called the combined probability of inclusion (CPI). The left glove was completely inconclusive, the expert testified. From the right glove the likelihood of selecting a random person with the same DNA profile as that in the gloves was 1 in 110 for the Hispanic population (other figures were presented for the African-American and Caucasian populations).

The jury was not compelled, with nine members holding out, leading to a mistrial.

The DNA analyst cut open the gloves and took more swabs. At the second trial, the right glove became “four to five times stronger” (1 in 450) for Hispanics. Additionally, the left glove then became inclusionary, to a 1-in-65 random chance of another Hispanic matching the evidence. The district attorney told the jury that the evidence was damning: that Dominguez was “16 of 16 markers” in the right glove and “14 of 16 markers” in the left glove, because of the mixture.

“It’s what the DNA tells you,” the prosecutor said at the time.

The jury agreed, and convicted Dominguez after more than three days of deliberation.

The DNA analyst in San Diego was Shawn Montpetit. (Montpetit did not respond to a Forensic Magazine request for comment.)


The court ruling found Montpetit “could (and did) ignore inconsistencies between certain markers in the known reference sample of the suspect and the corresponding markers in the questioned item.”

“The fact that Montpetit’s testimony was based on an analysis that was conducted using methods that were thereafter rejected by both the relevant scientific community and the SDPD Crime Lab was ‘favorable’ evidence that should have been disclosed to petitioner,” the judge wrote. “In his trial testimony on April 5 and 6, 2013, Montpetit—as a highly qualified forensic scientist, presenting ‘CSI-like’ expert testimony—testified to opinions and conclusions that, when he gave them, were invalid under the procedures then in place.”

Dominguez’s appeal based on the DNA mixture methodology commenced in 2015. The legal proceedings ultimately led to a new writ of habeas corpus—and potentially a new trial, according to court documents.


Suzanna Ryan, a Carlsbad-based DNA expert, provided testimony for the defense. She found that the San Diego analysis presented in 2011 did not meet its own protocols established just days before the second trial—and did not meet those established in 2010 by the Scientific Working Group on DNA Analysis Methods, or SWGDAM.

Ryan told Forensic Magazine that the DNA-mixture interpretation from less than a decade ago could mean that trials just a few years later could be tilted in the opposite direction, based on the best science now available.

“The thing is, is that there are cases like this all over the country,” she said. “If you were convicted of a crime pre-2010 (or even later in many jurisdictions) and the case involved a complex mixture, there is a very good chance the mixture would no longer be interpreted the same way.”

Prosecutors have realized that the CPI problem may go beyond population database errors, and instead go all the way from including a defendant as a contributor to inconclusive results, based on new SWGDAM standards, Ryan said. Texas is one state that has noticed potential problems, and attempted to remedy where misleading testimony may have swayed a case, she added.

“The Texas Forensic Science Commission is now proactively soliciting cases for review that involved the older mixture interpretation methods,” Ryan said. “They are the only state, that I am aware of, that is making an effort to address the mixture interpretation issues.”

The California appellate judge wrote that the trial’s DNA evidence had been unfair to the accused, no matter his ultimate innocence or guilty.

“Every defendant—including gang members—is entitled to a trial in which the scales were not weighted on one side by invalid scientific evidence,” the court ruled.

Ryan said that Dominguez and his attorneys could have new evidence at their disposal. The San Diego lab re-amplified two of the glove samples—and has used STRmix, a probabilistic genotyping software now gaining wide acceptance, to interpret the mixture.

Matthew Speredelozzi, the defense attorney, said re-swabbing or recollecting samples may not be part of a re-trial. Instead, it could be reinterpreting the data that's already within those gloves.

"Retesting of the biological material may not even be ncessary - the problem with the DNA was not the profiles," he said. "There was no issue with contamination that we know of, or profiles that were generated. Once you have those profiles, it's the interpretation of those profiles... that the habeas was based on."