LONDON — The history of science is littered with serendipitous findings that turn out to have important bearings on often completely unrelated areas of work.

The latest example of this phenomenon comes from a study which is helping to pinpoint the timing and place of the origin of the human race. Unexpectedly, this work has shown that a test for male or female sex, which is often used in forensic science, is less accurate than it was thought to be.

Chris Tyler-Smith, assistant director of the Cancer Research Campaign's Chromosome Molecular Biology Group in the department of biochemistry at the University of Oxford, in the U.K., and his colleagues Fabricio Santos and Arpita Pandya were examining the variability of DNA on the Y chromosome. Their study was part of an attempt to verify the finding that humans emerged in Africa as recently as 150,000 years ago.

This conclusion had been drawn several years earlier by other groups of researchers as a result of analyzing mitochondrial DNA, which is passed from a mother to her children.

One way of carrying out an independent check on this result would be to study the evolution of the DNA on the Y chromosome, which is passed from father to son. Initial results suggest the Y chromosome, too, had a relatively recent origin in Africa.

In the course of this study, Tyler-Smith and his colleagues were studying polymorphisms in the Y chromosomes of 350 DNA samples from men around the world. Their samples were drawn from every inhabited continent.

Using hybridization techniques and PCR tests, they looked for the presence of 37 different loci normally found on the Y chromosome. To their surprise, they found that two samples out of the 24 from Sri Lanka — 0.6 percent of the entire sample — lacked the Y chromosome copy of the amelogenin gene.

Tyler-Smith and his colleagues report their results in a letter to the February issue of Nature Genetics, titled "Reliability of DNA-based sex tests."

The amelogenin gene encodes a protein involved in tooth development. More significantly, however, it is also the gene used to help determine the sex of the DNA present in forensic samples. Both the Y chromosome and its counterpart X chromosome in males bear homologous, but not identical, amelogenin genes. These are called AMELY and AMELX.

Because the genes are homologous, the same PCR primers can be used to amplify them. But because the genes are not identical, they give rise to protein products of slightly different sizes, making it possible to distinguish whether only AMELX is present, or both AMELY and AMELX.

One advantage of this test is that it has its own built-in control: If amplification is successful, then it will always be possible to detect at least the AMELX gene product.

Tyler-Smith told BioWorld International, "The problem is that we have now found that a small proportion of men lack the Y copy of the amelogenin gene as a result of a natural polymorphism. The danger is that if people do not take this into account, they would score those males as females when using this test. If you had a forensic sample and wanted to know the sex but got the wrong answer, this could lead to a great deal of investigative time being wasted."

To get around this difficulty, Tyler-Smith and his colleagues suggested adding the appropriate primers for the SRY gene, which is directly involved in the determination of male sex. "This should make it possible to diagnose the correct sex with a much higher degree of reliability," Tyler-Smith said.

The group is continuing its analysis of the natural polymorphisms of the Y chromosome. Tyler-Smith concluded, "This finding, to me, is an example of how science works. We were not setting out to test or improve an assay for sex determination, but this shows that basic science does feed into important areas of scientific application."