By RAYYAN JOKHAI
For The News-Letter
The biology of sexual orientation has been one of the most complex and politically charged mysteries in human genetics. Researchers at the David Geffen School of Medicine of the University of California, Los Angeles (UCLA) have developed an algorithm that examines just nine regions of the human genome to identify associations between homosexuality and markers found on and around DNA.
The study examined the DNA in 47 pairs of identical male twins. Within each pair of twins, one brother was homosexual. The existence of twin pairs in which one is homosexual and the other is not can offer strong evidence that something other than DNA alone influences sexual orientation.
The researchers focused on identifying differences in DNA methylation, a type of modification to DNA that doesn’t affect the sequence, which stays the same for the identical twins. Epigenetic modifications like methylation generally occur due to environmental pressures and according to when and how strongly a particular gene is expressed. By using identical, twins the researchers could control for any differences in the genetic code and focus on the modifications to DNA that developed over time due to environmental factors.
While differences in genetic code were eliminated by examining the DNA of identical twins, identical twins still tend to have similar DNA methylation patterns, rendering the process of identifying potential epigenetic markers for sexual orientation challenging.
“A challenge was that because we studied twins, their DNA methylation patterns were highly correlated,” Tuck Ngun, the first author on the study, told the American Society of Human Genetics.
Even after some initial analysis, the researchers were left with over 400,000 data points to sort through. The innovation in their approach came through the research team’s development of a machine learning algorithm called FuzzyForest. The algorithm could quickly and independently sort through methylation patterns in the regions of DNA examined in order to identify particular molecular differences in DNA. Using the algorithm and the particular patterns it identified, Ngun and his team were able to predict study participants’ sexual orientation with 70 percent accuracy.
“To our knowledge, this is the first example of a predictive model for sexual orientation based on molecular markers,” Ngun told the American Society of Human Genetics.
The research findings have stirred conversation across the fields of genetics, biology, psychology and pharmacology. While many researchers appreciate the potential knowledge the study can add to the field, many are still concerned with the fact that the study’s results are preliminary. Additionally, some researchers have expressed concerns with the methods used for the study, citing the fact that the genes examined in the study were not all associated with brain development.
The results were presented earlier this month at the American Society of Human Genetics Annual Meeting in Baltimore and have not been published yet.