Box 1. Expert opinion

The following scientists have all previously testified in court cases involving HIV molecular evidence. They assess the new data.

"This study is an impressive statistical analysis. It shows clearly that the hypothesis of deliberately injecting children with HIV in 1998 should be rejected."

Philippe Lemey, expert on HIV evolution, Rega Institute for Medical Research, Belgium.

"This is exactly the kind of objective phylogenetic analysis needed in this case. The results clearly show that the health workers were not responsible for the introduction of these HIV strains."

David Hillis, expert on viral phylogenies, the University of Texas, Austin.

"This kind of analysis has been approved by courts around the world. This is a case of [hospital] infection with multiple, independent sources, a pattern most easily explained by sloppy or inappropriate practices at the hospital."

Fernando González Candelas, evolutionary geneticist, the University of Valencia, Spain.

"The existing epidemiological data are already enough to demonstrate that the accused medical staff cannot be the source of the contamination. De Oliveira's analysis is completely independent, and yields the same conclusion. The court cannot pretend to be impartial if it refuses to hear any competent scientist from abroad."

Michel Milinkovitch, evolutionary geneticist, the Free University of Brussels, Belgium.

"They have used state-of-the-art methods to estimate divergence and dates of events in this outbreak. The analysis shows compelling evidence that the outbreak had started before the accused could have started it."

Thomas Leitner, expert in HIV evolution, Los Alamos National Laboratory, New Mexico.