The DNA evidence used to link Mark Grant to the murder of 13-year-old Candace Derksen was not properly analyzed, according to a genetics professor.

The defence called on Dr. Bruce Budowle, of the University of North Texas Health Science Centre, who testified in the Grant retrial Monday.

Grant was convicted of second-degree murder in 2011, but the conviction was overturned two years later when an Appeal Court ruled the trial judge erred in not allowing Grant's defence to present evidence that pointed to another possible killer.

It was DNA evidence on a piece of twine tested in 1993 that led to Grant’s arrest in 2007.

Budowle, who has developed forensics validation methods for the FBI, told the court he has serious concerns about the analysis of Grant’s DNA done by Molecular World – the lab where Winnipeg police sent DNA evidence collected from twine used to bind Derksen’s wrists.

READ MORE: Defence builds case of ‘unknown’ third party suspect in re-trial of 1985 murdered teen

The lab ran three tests on the evidence. The first two tests did not produce usable results, but the third test was used to match Grant’s DNA to the DNA found on the twine.

Budowle said the analysis of the third test was “scientifically corrupt” and amounted to a “violation of DNA science.”

Budowle said DNA markers found in the third test could not have come from either Grant or Derksen. He explained that he believes the DNA used in the third test was not the same as the DNA used in the first two tests.

He said the results of the third test should have been considered inconclusive, and he questions the need to conduct the third test in the first place.

“It just doesn’t make sense,” he said.

Grant's defence has also raised questions about how police handled evidence and argued it could have been contaminated.

READ MORE: Court listens to how DNA was tested on day 5 of murder re-trial