A groundbreaking analysis of Neanderthal DNA that suggested they interbred with humans was based on samples contaminated with human DNA, a new study suggests.

The study1, published on 28 August in PLoS Genetics, reanalysed about one million base pairs of fossilized Neanderthal DNA that had been analysed in a paper published last November in Nature2. The Nature paper and a paper in Science3 published the same week on 65,000 base pairs were the first reports on Neanderthal nuclear DNA.

But around 80% of the sequences in the Nature paper are modern human DNA, not Neanderthal, claims Jeffrey Wall, an evolutionary geneticist at the University of California, San Francisco, who led the PLoS Genetics study. This indicates that human genetic material was somehow introduced into the samples. This known risk is increased by the closeness of the two species — the 3-billion-base-pair genomes of a human and a Neanderthal differ by less than 0.5%.

The results in the Nature paper suggested that there was interbreeding among Neanderthals and humans in their common European home ground before Neanderthals became extinct 30,000 years ago. The Science article found no genetic evidence of interbreeding.

Neanderthal DNA has been reanalysed, leading to suggestions of human DNA contamination. Credit: J. TRUEBA/MSF/SPL

Svante Pääbo, senior author of the Nature paper, concedes that his group at the Max Planck Institute for Evolutionary Anthropology in Leipzig, Germany, had problems with contamination. These prompted him to change laboratory procedures and to add controls late in 2006, after the paper was published. “I agree with [Wall's] analysis,” Pääbo says. “Their observations are formally correct.”

Pääbo's co-author Michael Egholm, who is research vice-president at 454 Life Sciences in Branford, Connecticut, adds: “There is no denying contamination. It was one of the dangers of doing this.” But ongoing analysis indicates that human contamination in their study was just 30%, Egholm adds.

There had been intense debate over the contrasting results in the Nature and Science papers, which analysed the same 38,000-year-old Neanderthal bones from Croatia using different sequencing methods. Pääbo's group used 454's rapid 'direct sequencing' approach, whereas the Science team, led by Edward Rubin of the Joint Genome Institute in Walnut Creek, California, used a traditional method using cloned DNA and bacteria to generate the base pairs.

The studies gave different estimates for the time Neanderthals diverged from humans — the Science article pegged it at 706,000 years ago, whereas the Nature paper set it at 516,000 years ago. Wall's study confirms the 706,000-year divergence date. The probable human DNA contamination led to the more recent date and may have led to the suggestion of later interbreeding, Wall says.

The Nature paper also found more similarities between genetic variations called SNPs (single nucleotide polymorphisms) in Neanderthal and human DNA than the Science paper, even after allowing for the Nature group's larger number of base pairs. Pääbo's team reported that about 30% of the SNPs in the Neanderthal DNA are derived — that is, the mutations occurred — in today's humans.

Pääbo acknowledges there is “a potential problem” with the presence of these human SNPs in the Neanderthal sequence. These same discrepancies were noted by Rubin's group. “We had concerns,” says Rubin. “We suspected some of the issues raised by Wall.”

Both Pääbo's and Rubin's groups expect to publish further Neanderthal sequences from other specimens that each group is studying. Pääbo and Egholm say their analysis will address the anomalies in their Nature paper.