A few days ago I complained about the web page for the RADC, which at that time looked like this:
I objected because I thought the page clearly implied that they could predict risk for addiction from DNA. I had just finished an extensive review of a similar effort by the same PI at her previous position at VCU, which showed unequivocally that they could not. I titled my post harshly: “An Unethical Research Program at Rutgers.”
In response (I assume) to my post, the web page has changed. Here it is today:
The phrase “DNA Discovery” has been removed from the title. Good for them, although I note that the worst part, “What’s in your genes?” remains. To be clear about my objection: telling people that their phenotypes are “in their genes” is never a very good idea, even if the genetic indicators in question are valid predictors. But that isn’t the main point here. The problem here is that these very investigators have just finished amassing clear and convincing evidence that they can’t diagnose anything significant about addiction from genetic data. They don’t have a polygenic score that predicts as much as 1% of the variance. So yes, the phrase “in your genes” is clumsy and essentialist, but in this case it is also just plain wrong. It is deceptive. They are saying they can do something they can’t do.
The text that follows emphasizes that the prediction platform they have developed combines genetic, behavioral and environmental information. Fine, but how does that work? If it is like what they developed at VCU, it consists of a set of ordinary self-report temperament questionnaires, which are then described as “genetically influenced”. The evidence for “genetic influence” presumably comes from twin studies, not the participant’s DNA. See my paper for more details.
So the best reading of the program is that they aren’t actually telling anyone what is in their genes, they are just assessing behavioral phenotypes like anyone else, and then declaring those phenotypes as “genetically influenced” based on the first law of behavior genetics. This is less harmful to the participants, but it is still obviously false advertising.
A word about vitriol. I heard from the PI, whom I have known for years, after the first post. Unsurprisingly she was not very happy. She did not offer any reasons why my assessment of the research program was incorrect, but wondered why I felt it was necessary to be so vitriolic.
What am I supposed to do? This is more than a simple disagreement about scientific conclusions. The web page in question is designed to recruit research participants, who are being systematically misled about the nature of the research and what it can accomplish. But like I said the first time, it is not my place to police the ethics of other people’s research— that is between them and their IRB.
What really troubles me is that much of modern behavior genetics seems to have abandoned any commitment to empirical results. The idea that addiction (or pretty much any other behavioral phenotype) is “in your genes” has become so baked in that nothing can dissuade investigators from pursuing it. What would it take? How many times are they going to report essentially null results from genomic data before they stop making claims like that?
I am proud to be a behavior geneticist. I am a former president of BGA and I love the field and its tradition. But what I love is its commitment to science and scholarship. Advertising research based on obviously insupportable premises is damaging to participants, but it is also damaging to the enterprise responsible scientists are trying to conduct. If that is vitriol so be it.