A pocketknife used to kill Mabel Greineder contained miniscule amounts of material evidence that shared striking characteristics with her husband Dirk Greineder's DNA, a genetics lab scientist testified Thursday.

The testimony marks a potential blow to Greineder, a renowned Boston-area allergist accused of killing his wife in October 1999 so he could pursue a secret life of pornography and trysts with prostitutes.

Under cross-examination, however, defense attorney Martin Murphy picked apart the methodology of the geneticist, Robin Cotton, arguing that the laboratory, Cellmark Diagnostics, would have reached entirely different results had the scientists used methodology sanctioned by the FBI. Cotton acknowledged that the federal government uses a much higher threshold for positive tests of genetic material than Cellmark.

Throughout her direct testimony, Cotton reeled off a number of figures and calculations that hurt Greineder's chances of avoiding a first-degree murder conviction for slaying his wife in a placid park near the couple's suburban home.

The scientist explained that, depending upon the quality of evidence, Cellmark's DNA testing can analyze up to 13 different "locations" in the genetic code of a sample. A numerical value is affixed to each location depending upon its character, and those numbers can then be compared to the values found in other samples.

"The data from the knife indicates that there is a primary source and a secondary source and that the primary source is from a female and the secondary source is from a male," she said. The secondary source points to characteristics that match Greineder's DNA, she said.

Once the level of similarity between two samples is determined, Cotton told the jury, the next step is to figure out the chance that a random person would have the same DNA characteristics.

Defense lawyer Martin Murpy cross-examines the witness as the Greineder children look on (Court TV photo)

Only one in 2,200 white males would possess the genetic characteristics found on a small pocketknife allegedly used to slit Mabel Greineder's throat, she testified. And only one in 680,000 would match the DNA sample lifted from a bloody glove found with the weapon in a storm drain near the crime scene.

The victim was killed while on a walk with her husband in the Morse's Pond Recreation Area, a wooded park in the affluent suburb of Wellesley, where the couple with their three children had lived for over 30 years.

Greineder says he and his wife were walking their dog when she hurt her back. He claims they decided to separate so that he could continue exercising their pet, and that an unknown assailant struck in his absence.

But authorities maintain that this story is just part of an "elaborate plot gone awry" by the doctor to murder his wife and make it look like the work of a serial killer. Prosecutor Richard Grundy says Greineder was secretly obsessed with pornography and prostitutes, and wanted to pursue these passions more freely.

The defense theory of an unknown killer was bolstered by testimony concerning DNA found on a glove the victim was wearing. The scientist told the jury that cellular material on the glove could not be traced to either Mabel or Dirk Greineder, a point that defense attorney Martin Murphy emphasized during cross-examination.

"That sample when it was analyzed revealed the presence of DNA of at least 3 people, is that correct?" he asked.

"Yes," Cotton firmly replied.

"It could not have come from (either Greineder)?" he pressed again.

"That's correct," the witness stated.

The defense lawyer asked a grueling series of question about the treatment of the DNA samples by Cellmark — from the manner in which they arrived at the lab to the ways in which they were analyzed.

Much of this line of inquiry seemed routine and tedious until Murphy started asking about how much material at each "location" on the genome in a DNA sample was required for the lab to report a positive test in that "location."

Cellmark Diagnostics changed its standard during the spring of 2000 in this regard, after only some of the evidence in the case had been given to them to analyze. Some items were thus treated with looser standards for reporting the presence of genetic material.

"If the data had been sent to you later, your analysis would have resulted in a different set of charges as were displayed during direct examination in respect to the samples that were sent to you in rounds one and two?" Murphy asked.

"Yes," Cotton replied.

"All those numbers would disappear?" he pressed, after ticking off several values that would not have survived the stricter standard.

"Yes they would," she acknowledged.

But the most stunning admission Cotton made involved the difference between her lab's standards and the FBI's methods. Murphy got her to acknowledge that the federal government uses a much higher threshold for positive tests of genetic material than Cellmark.

"If the sample from the work glove had been sent to the FBI rather than Cellmark, the results would have been different?" the lawyer asked.

"Yes. They wouldn't have gotten anything," Cotton acknowledged.

"They would have recorded non-detects across the board?" he said again.

"Yes they would have," she said.

Testimony resumes Friday morning at 9 a.m. The trial is being broadcast on Court TV.