BEIJING — China's top veterinary official on Friday rejected as groundless a study that reported the emerging dominance of a bird flu strain in Asia, saying it had no basis in science.

Jia Youling, director of the Agriculture Ministry's veterinary bureau, also said 20 virus samples had been sent to a World Health Organization lab in the United States. It was a key development for the health body, which has criticized the ministry for withholding H5N1 samples since 2004.

The latest study, published Oct. 31 in an international scientific journal, said the strain - called H5N1 Fujian-like - was found in almost all poultry outbreaks and some human cases in southern China, and was now also the primary version in Hong Kong, Laos, Malaysia and Thailand.

The findings angered Chinese officials, who have been trying to overcome a reputation of foot-dragging in co-operating on investigating emerging diseases such as bird flu and SARS. Since the study's release, China has issued unusually strongly worded responses disputing the results.

"There is no such new Fujian-like virus variant at all. It is utterly groundless to assert that the outbreak of bird flu in Southeast Asian countries was caused by AI (avian influenza) and that there would be a new outbreak wave in the world," Jia said at a news conference.

"The data cited in the article were unauthentic and the research methodology was not based on science," he said. "Therefore their arguments were not tenable and totally against the facts."

The latest findings were released in the Proceedings of the National Academy of Sciences, published in the United States following a study conducted by scientists in Hong Kong and the United States.

The scientists include respected bird flu experts Guan Yi and Malik Peiris of the University of Hong Kong and Robert Webster of St. Jude's Children's Research Hospital in Memphis, Tenn.

Guan did not immediately respond to e-mails or phone messages on Friday. In an e-mail earlier this week, following earlier Chinese criticism, he stood by the findings and said the data "was subject to scientific peer review and published in an international journal of high repute."

The study charted the spread of the Fujian-like strain by testing geese and ducks found in markets in six southern Chinese provinces beginning in June 2005.

Over the course of the year, the strain became more pervasive, the study said. Among the 108 samples taken from Chinese poultry in April and June 2006, 103 - or 95 per cent - were infected with the Fujian-like strain, it said.

The researchers said it was unclear how the strain had emerged or spread so widely. They also said China's current poultry vaccines appeared to be ineffective against the strain.

Jia questioned aspects of the research including the sample size, how samples were collected and the accuracy of the conclusions. He said they did not follow proper procedures for vaccine testing, unlike China, which has tested millions of samples.

"I think it is irresponsible for them to say our vaccines are not effective simply because of results from experiments with a few samples," he said. "It is unacceptable to us."

Last week, WHO urged the Agriculture Ministry to share its samples, a key step in helping health experts track the diseases and ultimately develop vaccines.

But Jia said "the virus samples have been delivered to the WHO. This demonstrates China's readiness and willingness to co-operate."

Henk Bekedam, the WHO's representative in China, said he was encouraged by Friday's developments and that the agency looked forward to receiving the samples from 2004-05.

"I'm very optimistic," Bekedam said. He said WHO also will request samples from 2006.

"We still strongly believe it's important to get a proper picture," Bekedam said. "China is very big. The more information you get, the better it is. We are dealing with a new challenge, a new threat."

New virus strains emerge regularly, but health experts need to know when one becomes dominant so they can try to develop a vaccine.

Concerns are particularly strong that the bird flu virus could mutate into a strain easily passed between humans, and spark a pandemic. Most human infections have been traced to contact with sick birds.

The H5N1 virus has killed at least 152 people worldwide since 2003, according to the WHO.

Jia's statement said that, from January to the beginning of November this year, there had been 10 bird flu outbreaks in seven provinces, killing 47,000 birds.

It said authorities destroyed 2.94 million birds in order to contain the outbreaks.