Studies into sexuality have always been the rage with scientists, with the nature versus nurture debate at the heart of our attitudes to sexuality.
Recently, a team of post-doctoral scientists have published findings from a study in which they claim to have found a method for correctly predicting someone’s sexuality, 70% of the time, by simply studying regions of their genome.
A team from the University of California, Los Angeles, headed up by Tuck Ngyun presented the results to an audience at the annual conference for the American Society of Human Genetics last week.
The team conducted a study into 47 pairs of identical male twins. 37 of these pairs had one gay and one straight twin, and 10 pairs had twins who were both gay. The team scanned 140,000 regions of the particpants’ genomes and whittled the areas of interest down to a potential 6,000. They then built computer models which investigated methylated areas of DNA, which are ‘switched on’. The best algorithm generated by the team’s computer models predicted the sexuality of the participant by looking at 5 methylated points of DNA. This was accurate in 67% of the cases.
A press release based on the abstract of the study was sent out by the organisers of the conference to media, and within hours, the story of a definitive ‘gay gene’ was being worked into print the world over.
At the conference, however, Tuck Ngyun and his team were ridiculed for the weakness of their results, and asked if it would be possible to reproduce the study and consistently predict the sexuality of more people.
Due to a lack of funding, the team ran the study with a tiny sample of participants, which in itself was inadvisable. The findings can’t be generalised, which sensationalist media have not taken into account. In scientific studies, a result of 67% in a tiny pool of participants is statistically insignificant.
“The story of a definitive ‘gay gene’ was being worked into print the world over”
The team also split the participants into a ‘testing’ set and a ‘training’ set. Information from the ‘training’ set was used to generate the algorithms, and then tested on the second group. So a study that was already statistically insignificant suddenly becomes doubly so.
Similarly, the best algorithm generated by the team’s computers only found 5 methylation markers out of 6,000 regions of interest, which could have been found by chance alone. 5 markers out of a possible 6,000 once again do not prove anything.
Other scientists in the field of epigenetics have pointed out that although the methylated genes are thought to be involved in immune genes which have been linked to sexual attraction, the correlation between the methylation of these genes and its influence on their function cannot be proven by this study.
Methylation of DNA can occur at any point in a person’s lifetime, and can be caused by internal factors (e.g. genetic factors) or external factors in the environment. So, even if the results from this study could be counted upon to accurately predict sexuality, it can’t be used to predict if a child will grow up to be gay on methylation markers alone, because there are too many factors involved.
To further delineate the conclusions of the study in the question; it only focuses on homosexual men. Regardless of further replications of the study and similar studies – a ‘homosexual gene’ can’t be coined unless it informs female sexuality, too. What about homosexual women?
And that’s before even taking into account the spectrum of sexuality that exists – if genes inform homosexuality and heterosexuality, then are there ‘asexuality’ and ‘bisexuality’ genes too? Or is research of this kind a Pandora’s box, and better left unopened?
The ethical implications of defining biological markers for sexuality are varied and unnerving in the extreme.
If it were possible to predict a child’s sexuality, how would that inform their upbringing? In a world where clinics for ‘conversion therapies’ exist, so that homophobic parents subject their gay offspring to therapies which claim to correct their sexuality, would a biological basis for predicting homosexuality do more harm than good?
Although it would be great to prove that some of us are just ‘born this way’, such technologies could give homophobic parents the power to abort foetuses with the biological markers.
Looking at the wider picture, how would regimes that condemn homosexuality utilise such markers? In countries where homosexuality carries a death penalty, a positive ‘gay gene test’ could spell a death sentence.
We’re not anywhere near isolating a ‘gay gene’. But maybe that’s a good thing.