Researchers search for genes behind intelligence, find almost nothing | Ars Technica

Three possible genes, all with minuscule effects on brain power.

Flickr user Charles Strebor

To what degree do genes determine how intelligent we are? It’s a question with a lot of social implications, given the potential for people to get lumped into categories based on their inheritance—or for them to make a self-fulfilling prophecy out of it.

But so far, at least, it’s a question that science has been unable to answer. When scientists have focused on specific genes involved in the nervous system’s growth and function, they’ve found a few hints of associations, but those haven’t held up in large population studies. Large population studies that scanned the entire genome have also come up empty.

Now, we may have a good idea why. A massive and thorough genetic study has looked into the question and, despite all the effort involved, has only come up with three genes, none of which have a very large effect on cognitive abilities.

The work was done by a collaboration of nearly 60 scientists. Rather than either looking at a specific set of candidate genes or scanning the entire genome, they took a hybrid approach. Reasoning that intelligence is likely correlated with academic achievement, they first did a scan of the genome, looking for base differences that were associated with educational attainment. This approach, called a "genome wide association study," relies on tens of thousands of individual bases that normally vary within human populations. To make sure that whatever came out was likely to be significant, the work relied on a panel of more than 100,000 participants.

This left them with over 900 individual sites in the genome. Many of these, however, were physically close to each other, and thus were almost certain to represent the effect of a single gene. Once this was accounted for, the authors were left with 69 separate sites in the genome. Those 69 locations were then tested in a separate population of over 24,000 individuals, all of whom had been given tests that measured their cognitive abilities.

To avoid some of the problems with the earlier work (which hadn’t been replicated), the authors had pre-registered this experimental design with an open science hosting site. They also had two teams independently repeat and verify the analysis. So, from an experimental perspective, the results are likely to be solid.

But from an informational perspective, they were rather weak. To begin with, there was no significant correlation between educational achievement and intelligence test scores associated with these sites in the genome. (In fact, there was a weak negative correlation, but it went away if you discarded a single extreme outlier.) In terms of intelligence, the majority of the 69 sites the authors looked at were within the 95 percent confidence interval, meaning they weren’t significantly different from random chance. Only three individual sites appear to stand out above the statistical noise.

The effects size, which measures how large the influence of the genetic regions is, was also quite small. Human height, notable as being difficult to associate with a genetic influence, has genes with an effect size 20 times larger. Even if someone had two copies of the variants associated with intelligence at one of the genes, their IQ would get a boost of about half a point compared to someone who lacked them.

Then there’s the three genes that are closest to these sites in the genome. The authors don’t mention much about them, and it’s easy to see why. None of them are clearly involved in nervous system development. Only one of them looks like it might have a biological function that we know about (it’s related to enzymes that modify how DNA is stored in the cell.) The second simply has a domain that typically is used to allow two proteins to interact, while the third may not even exist—its record was discontinued after a reanalysis of genome data.

The authors say that to find anything else, we’ll have to go to samples that are larger still, which would involve a staggering number of genetic tests. And, if these (and past) results are any guide, then they’re not likely to come up with anything very exciting.

All of which suggests that intelligence isn’t a matter of a handful of genes. Sure, there are some genes that, when damaged, have a catastrophic effect on cognitive abilities. But, assuming you avoid these, it seems that your intelligence is likely to be the product of a huge collection of minor genetic effects, combined with a very large helping of your environment.

PNAS, 2014. DOI: 10.1073/pnas.1404623111 (About DOIs).

Reader comments

John Timmer / John became Ars Technica’s science editor in 2007 after spending 15 years doing biology research at places like Berkeley and Cornell.

@j_timmer on Twitter

Evernote helps you remember everything and get organized effortlessly. Download Evernote.
Advertisements

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s