AUGUSTA, Maine — A faulty FBI software program used by the Maine State Police Crime Laboratory to determine the probability that DNA collected at the scene of a crime matches a certain individual has the potential to affect thousands of criminal convictions in Maine.

“This could be a huge issue, especially in cases where DNA evidence was a significant factor in the case,” Augusta defense attorney Walter McKee said recently in an email. “[The Maine Association of Criminal Defense Attorneys] has put out the word to its members about this incredible development to determine past cases where this could be an issue. The full impact of just how many cases this will affect won’t really be known for some time.”

One of the cases that could be affected is that of convicted murderer Jay A. Mercier, 59, of Industry. He is serving a 70-year sentence at the Maine State Prison in Warren for the July 5, 1980, slaying of 20-year-old Rita St. Peter of Anson. A jury convicted Mercier in September 2012, more than 30 years after St. Peter died.

Mercier told police he did not know St. Peter and did not take the stand in his own defense.

Andrew Benson, the assistant attorney general who prosecuted the case, said after the verdict was announced that he believed the DNA evidence played a pivotal role in the case.

“I think the critical piece of evidence in this case was the fact that Ms. St. Peter had obviously just had a sexual encounter because of the nature and the disarray of her clothing,” Benson said. “And the fact that Mr. Mercier’s DNA — his sperm — was not in her underwear, only in her [body].”

The evidence linked Mercier to the smear slides taken from St. Peter’s body in 1980 through a discarded cigarette butt he left on the ground in January 2010 on the road outside his home during an interview with the detective assigned to cold cases. DNA taken from the cigarette matched DNA taken from St. Peter’s body in 1980.

Crime lab technician Kathy MacMillan testified at the trial that the probability of the DNA belonging to someone other than Mercier was “one in trillions.”

Mercier’s attorney, David Paris of Bath, said Friday he has asked the crime lab to recalculate the DNA probability in the case. Paris is handling Mercier’s post-conviction review but was not involved in his trial.

The retest is not expected to rule out Mercier’s DNA, but the probability that it might be someone else with the same DNA profile is expected to change, he said.

“I don’t know what the results may yield,” Paris said.

Lisa Marchese, head of the criminal division in the attorney general’s office, said Friday she’s “not expecting many, if any, convictions will be impacted.”

Penobscot County District Attorney R. Christopher Almy said it was too early to tell how many if any cases may be challenged because of the use of the faulty probability data.

The FBI in May notified labs around the country of the errors in the data used to make the statistical DNA calculations between June 1, 2010, and June 1, 2015, Lt. William Harwood, head of the crime lab, said in an Oct. 10 letter addressed to the “Criminal Justice Community.”

The FBI told crime labs that the “problem stemmed from ‘clerical mistakes in transcriptions of the genotypes and to limitation of the old technology and software,’” the Washington Post reported May 29.

“It would be impossible for us to know what, if any impact, statistical estimate variations would have had on the adjudication of a past case,” Harwood said in the letter. “For example, in one experiment with old and new data a random match probability changed from a genetic sequence being estimated to be found in 1 in every 236 quadrillion people to 1 in every 219 quadrillion people.”

Harwood said the statistical impact on cases that included the results of tests of low-level partial DNA sequences, often found in touch DNA, could be more significant. Touch DNA is made up of skin cells people shed when they touch an object.

“In one experiment with low-level DNA, we attempted to produce a scenario that would create a more significant difference in the results from using old and new data,” Harwood said in the letter. “In this case we assumed a two locus DNA profile. The random match probability changed from a sequence found in one in every 32,960 people to one in every 28,570.”

Harwood said recently in a telephone interview that the crime lab performed about 8,000 tests on DNA between June 1, 2005, and June 1, 2015. Some results showed there was no match while others found the sample was not large enough to yield results. Other results turned over to prosecutors did not result in charges being filed, he said.

Harwood said he did not know how many tests might have led to convictions.

“The FBI has told us that the biggest swing we could see is two-fold either way with touch DNA,” Harwood said. “So, if the probability had originally been one in every 500, it could go to one in every 250 or one in every 1,000.”

Harwood said he expected the lab would be able to handle requests for recalculations with its current staff and within its current budget.

Defense attorneys and the Department of Corrections are working to notify as many people who were convicted during the time period when the faulty statistical model was in place, Marchese said.

“Our focus now is on getting people notified,” she said.

The Indigent Legal Defense Commission, which keeps a roster of lawyers who take court-appointed cases, and the Maine Association of Criminal Defense Attorneys have been asked to notify current and former clients whose cases might be affected, Marchese said. The DOC is working to let prisoners serving their sentences and people who are on probation know about the problem.

John Pelletier, executive director of the commission, said Friday that the commission would provide an attorney for people who could not afford to hire one to look into whether they should ask to have the DNA evidence presented in their cases re-tested.

“The next step would be to consider whether to file a motion for a new trial or a post conviction review,” he said. “At this point, we don’t think it would impact our budget.”