Y-chromosome profiles are male-linked genetic markers that have pointed detectives in the direction of serial killers who have remained at large for decades. However, the male-linked genetic information has also led to at least one erroneous identification and false conviction, in Taiwan.

Y profiles should be used in criminal investigation and in prosecutions—but need to be presented with particular caution in front of juries—according to a new study focusing on the use of the genetic profiles published in the online journal PLoS Genetics.

“Even under the optimistic assumption that the database has been sampled randomly in the relevant population, we show that database counts don’t help much,” write the two scientists. “The reason is that modern profiling kits with high profile mutation rates imply that almost all profiles are rare relative to typical database sizes.”

The two scientists performed computer simulations of a hypothetical population, and also assessed a half-dozen real Y-profile databases in Europe. (The two authors, Mikkel M. Andersen of Aalborg University and the University of Copenhagen in Denmark, and David J. Balding of the University of Melbourne and University College London, are forensic geneticists.) The software they used is an open-source program called MALAN (Male Lineage Analysis).

The computer model used the parameters and mutation rates captured by the view of three popular DNA profiling kits: Yfiler (17 loci), PowerPlex 23 (23 loci) and YFiler Plus (27 loci).

The statistical exercise involved a population of 7,365 that grew to more than a million over multiple generations. Their goal was to find how commonly an individual, nicknamed Q, shared his Y-chromosomal profile with other men in the three generations expected to be alive at the time.

Andersen and Balding then compared their data with the genetic variety in six Y-profile databases found in Europe, the largest containing more than 5,300 profiles in Central Europe, and the smallest containing just over 750 profiles in Southeastern Europe.

Their conclusion: Q cannot be identified with the same near-complete assurances provided by a full sample of nuclear DNA—but Y profiles could provide a major lead to developing investigations, they conclude.

“It should be clear from these considerations that a matching Y profile, taken alone, can never suffice to establish convincingly that Q is a source of the crime-scene DNA,” they write. “However, it remains very powerful evidence that can reduce the number of alternative sources from perhaps several millions, for a crime in a large city, down to just a few tens.”

Although the two scientists concede that forensic DNA experts are not entitled to make some of the legal judgments, “which are a matter for the court,” they do suggest a way to present the findings to judge and jury, explaining that it is not an exact match, and focusing on a 95 percent confidence intervals for a population frequency.

The Y profiles are especially of use when males assault females, to differentiate possible DNA mixtures, according to the authors.

Y-STR analysis has been vital to several key breakthroughs, most notably in the arrest and death penalty conviction of the Los Angeles serial killer known as the Grim Sleeper, Lonnie Franklin Jr. It was a partial match between the long-at-large killer and Franklin’s son that led detectives down the trail to catching him after decades. The killer known as China’s “Jack the Ripper” was also caught through Y profile analysis in August 2016, after killing at least 11 women and girls, and remaining at large for decades. Y-STR analysis brought authorities to 52-year-old Gao Chengyong’s doorstep, after a partial match to his incarcerated uncle. Both breakthroughs were made when Y matches were made, which gave investigators a lead that led to further evidence that solidified their cases; in the Grim Sleeper case, police posing as waiters took used utensils from Franklin, which proved he was the serial killer.

However, the first exoneration based on misleading Y-STR analysis, which happened in Taiwan, was reported and explained earlier this year in the journal Forensic Science International: Genetics. A third suspect in a suspected gang rape shared genetic information with two other suspects—and was assumed to have been included in a complex mixture, until the 2014 guilty conviction reversal.