The Japanese scientist accused of falsifying data in a widely heralded stem-cell research paper said Wednesday the results are valid despite mistakes in their presentation.

Haruko Obokata, 30, struggled to maintain her composure during a televised news conference packed with hundreds of reporters, but insisted she did not tamper with the data to fabricate results. She said she did not agree with retracting the research that appeared in the scientific journal Nature.

The phenomenon of ordinary cells transforming into stem cells under the influence of external stimulus was “confirmed many times for a fact,” Obokata said at the press conference in the western city of Osaka. She said she could replicate her results again if allowed back into the lab to do so.

She has contested allegations of research malpractice made by a panel of scientists at the government-funded Riken Center for Development Biology, where she is employed.

Obokata said Riken had advised her not to speak publicly about the controversy. The event Wednesday, which lasted for over two hours, was her first public appearance in weeks and was made against doctors’ orders after seeking hospital treatment Monday.

“I wasn’t able to sleep well and was feeling worse and worse,” she said. “I take this responsibility severely. I am deeply sorry.”

Obokata apologized repeatedly for having used the wrong images and having altered an image in a report on using a simple lab procedure to grow tissue for treating illnesses such as diabetes and Parkinson’s disease.

In an emailed statement, Riken reiterated Obokata’s right to an appeal of its findings.

“Riken will give due consideration to the appeal in accordance with our regulations,” it said.

Obokata attributed her mistakes to inexperience and her “limited abilities,” and appealed to fellow researchers to continue the work.

“The research was done accurately and the data exists to prove it so I would like to ask for your understanding that the research paper was not completed with a malicious intent.”

Obokata said she opposes retracting the paper published in Nature because that would give the international scientific community the impression that the whole project was invalid.

A lawyer who appeared with her went through a detailed explanation of the findings of Riken’s panel and concluded with a call for a review and longer evaluation of the research.

“The investigation was inadequate. It was carried out very quickly. It needs to be redone,” said the lawyer, Kazuhiko Murotani.

In announcing its findings, Riken faulted Obokata’s fellow researchers and senior staff for failing to fully verify the results in the paper submitted to Nature. But it held Obokata solely responsible for any fabrication or other mistakes in the data.

Obokata said she needs access to her computer in the laboratory and to her research notebooks to be able to prove she succeeded in turning cells from spleens of newborn mice into stem cells by exposing them to a more acidic environment than they are used to.

If replicated with human cells, using such cells to replace defective tissue might eliminate the risk of transplant rejection in patients suffering from various diseases.