The major blunder that sent detectives on the Gareth Williams inquiry down a blind alley for more than a year is the second embarrassment admitted this month by Britain's biggest private forensic science laboratory.
The latest error – described as "astonishing" and which led to a public apology from the company to the Williams family – has raised questions over systems and practices at LGC, which handles half a million samples a year.
Critics called for a public inquiry to try to discover if private providers were fit for purpose following the final closure today of the state-run Forensic Science Service. "This simply should not happen," said Professor Peter Gill, a founding father of DNA profiling in the 1980s. "Surely there is sufficient evidence to open a public inquiry because we are now getting regular reports of widespread system failure."
The mistake followed the discovery of a partial trace of a second person's DNA on the hand of Mr Williams. To rule out possible contamination from any of the investigators at the scene, those details were loaded manually onto a database.
But that entry was never checked and a simple typographic error meant no match showed up. It was only last week, when the files were checked, that scientists realised the correct trace matched the DNA of a Metropolitan Police scientist.
"We are sorry for any pain this error may have caused Mr Williams' family," said LGC in a statement.
The man who recorded the sample faces an internal inquiry but the ramifications for the company could be more serious. It said yesterday that it had checked back on four years of samples to rule out other errors.
Westminster Coroner Fiona Wilcox said: "There are wider concerns about whether LGC can be trusted to provide a proper level of forensic back-up to all investigations, remembering they are providing forensic DNA and analytical background to a lot of criminal cases."
The failure comes amid a continuing investigation by the forensic science regulator, Andrew Rennison, into another problem at an LGC laboratory in Teddington, southwest London. A contaminated sample led to a man being charged with rape in a city that he claimed he had never visited.
Adam Scott, from Exeter, Devon, was not considered a suspect but was charged after tests on a sample retrieved from the crime scene in Manchester revealed a match. The charge was dropped after LGC accepted that the samples were mixed.
The initial inquiry by Mr Rennison blamed human error and said 26,000 other samples had been checked and no further cases had been identified. Mr Scott's solicitor, Philippa Jefferies, said the latest case reinforced her calls for a public inquiry. "I'm concerned that two basic errors have happened now," she said.
The Home Office said it was "inappropriate" to comment because of the inquest and Met police inquiry.