Sunday, October 25, 2015

This Time for Sure! I Mean It! No, Really!

One story that got a lot of attention while I was getting ready to travel was an announcement of the latest "gay gene" study.  I first learned about it from a Queerty story linked on Facebook by several queer friends: "Researchers Say They May Have Found the Gay Gene For Real This Time."

The story, apparently drawn from a report at NBC News,  was about a study led by Tuck C. Ngun at "the David Geffen School of Medicine of the University of California, Los Angeles" (really! seriously! this story did not come from The Onion!), which was announced at a genetics conference, and promptly went viral, as these things generally do.  It doesn't appear that Ngun actually claimed to have put salt on the tail of the gay gene, as Queerty and other media said.  What Ngun actually said was something along the lines of "a predictive model for sexual orientation based on molecular markers."  Close enough, I guess.

On the other hand, the announcement drew a lot of criticism.  Even the Queerty article quoted some scientific critics.  Very soon a full demolition appeared at the Atlantic.  Ngun's study used a very small sample, then split that sample, and
As far as could be judged from the unpublished results presented in the talk, the team used their training set to build several models for classifying their twins, and eventually chose the one with the greatest accuracy when applied to the testing set. That’s a problem because in research like this, there has to be a strict firewall between the training and testing sets; the team broke that firewall by essentially using the testing set to optimize their algorithms.
... Ngun admitted that the study was underpowered. “The reality is that we had basically no funding,” he said. “The sample size was not what we wanted. But do I hold out for some impossible ideal or do I work with what I have? I chose the latter.” He also told Nature News that he plans to “replicate the study in a different group of twins and also determine whether the same marks are more common in gay men than in straight men in a large and diverse population.”
Great. Replication and verification are the cornerstones of science. But to replicate and verify, you need a sturdy preliminary finding upon which to build and expand—and that’s not the case here. It may seem like the noble choice to work with what you’ve got. But when what you’ve got are the makings of a fatally weak study, of the kind well known to cause problems in a field, it really is an option—perhaps the best option—to not do it at all. (The same could be said for journalists outside the conference choosing to cover the study based on a press release.)
Now, now -- expecting journalists to actually wait for research to be published, and to examine it critically, would spell the death of modern science journalism in the mass media, as we know it today.

I was amused by Ngun's protest -- better to do badly designed research than to do none at all!  It doesn't seem to have occurred to him that going public with such poor work would hurt his chances of doing the bigger study he hopes to do in the future.  Maybe he should bypass all the grinches among his colleagues who picked his baby to pieces, and try crowdfunding the next one.  GLBTQ media like Queerty would be glad to help.  The comments under that article are appalling; but then so are many of the comments under the Atlantic article, including a batch by a minister who asserts that sexual orientation is "as immutable as eye color and hand dominance," without any evidence to support that claim.

One other curious thing about the Queerty article, though.
Interestingly, after making the findings, Ngun, who is openly gay, decided to abandon the research out of fear that, if developed further, it could be used to screen fetuses or punish or persecute gay people.

“I just left the lab last week,” he said. “I don’t believe in the censoring of knowledge, but given the potential for misuse of the information, it just didn’t sit well with me.”
No link, no source given at all.  It's not mentioned in the NBC article that Queerty linked to.  So, did Ngun leave the lab before or after he announced his findings to the American Society of Human Genetics?  Before or after he defended his inadequate work against criticism?  Why did he only consider the "potential for misuse of the information" after he'd done the study?  It's not as if those concerns haven't been raised often before.

I found a probable source for the quotation from Ngun in a New Scientist article.  Another NS article pooh-poohed the concerns in equally familiar ways: Mankind must not be left ignorant of trivial and unnecessary matters, no matter what the human cost.

Then I found another article at the right-wing Daily Wire, written in a language I believe the author thinks to be English, which announced that
Not only had the study been misrepresented in a way that led many publications to announce that a "gay gene" had been found, it was done so all without the necessary authorization of the senior author and principal investigator of the research, Dr. Eric Vilain. What is more shocking is that the scientist who presented the research, Dr. Tuck Ngun, who is openly gay, was offended by his own research and decided to completely abandon the lab a week before the conference for fear of it shedding a negative light on homosexuality.
This, along with the rest of the piece, is fatuous.   No amount of care in presentation could prevent "many publications" from interpreting the study as evidence of a gay gene; the mass media, straight and gay, are extremely fond of biological determinism and will impose it on any story they cover.
In fact, the study was proving the opposite of what the public was led to believe: that there is no gay gene. Unfortunately, that reality was too harsh for the politically correct to accept, because that would mean that little boys who wear makeup are not genetically gay; they are still just little boys wearing makeup. It would diminish the need for a 'gay community' and threaten the validity of many males who claim that they are gay. 
I can't find any support for the claim that Eric Vilain had any objections to Ngun's presentation, but it appears that he too favors a genetic explanation for homosexuality: "The twin studies do not show that it is 100% genetic. They just demonstrate that there is a genetic influence."  So Vilain is also one of the "politically correct" that the Daily Wire denounces.  And the claim that Ngun's flawed research can already "predict" someone's sexual orientation is still being trumpeted, despite his critics.