There are now several companies offering to read your genes: 23andMe, Ancestry DNA, and Family Tree DNA are a few of the most prominent. You send them a saliva swab and cross their palm with silver, and back comes an analysis of your DNA.
There are plenty of grounds for skepticism about the new business of personal genomics. Some people are concerned about privacy and security: Your genetic data can be legally sold if it is first anonymized. We should also be worried about unscrupulous companies delivering “bad” news (like having the breast-cancer gene variant BRCA1) without free and thorough counseling advice. And some of these companies apply subtle emotional blackmail to sell their wares. Don’t you owe it to your kids to find out which genes you’ve given them? Don’t you owe it to yourself to have this precious knowledge of your “true being,” your molecular barcode of identity?
DNA-testing companies help perpetuate that latter dogma. “Welcome to you” says 23andMe’s saliva collection kit. “We all have our unique set of DNA that predetermine [sic] who we are,” says Futura Genetics’ website.
But in fact, your genes actually say very little about you—they’re not “what make you you.” To say this is wrong scientifically, philosophically, and arguably even ethically. Only when people start to realize this will they will see through the damaging rhetoric.
And that’s the real reason why, in the end, it might be a good thing that so many people are paying to get their genomes read—because it might finally shatter the myth of what genes say and do. Receiving a rather deflating and ambiguous list of boring traits and a summary of your “heritage” should do the trick.
Big promises, disappointing results
The marketing hyperbole of DNA-testing companies is echoing decades of misinformation from the genomics community. The Human Genome Project was sold to the public on the premise that the genome is “the book of life,” giving the impression that it is a kind of destiny written in molecular form.
But your genes are at best a rough sketch of potential yous, shaped and modified by outside influences. Genes can be turned on and off or ramped up and down by signals from the environment—and that’s before we even get into the way learning shapes personality traits and the brain.
“The possible misconception about how deterministic the genome sequence is will disappear with more information,” says geneticist Javier Herrero of University College London. “As more and more people get their genome sequenced, the banalisation of the genomic information will increase awareness.”
Some people have already attested to a “meh” response when they’ve had their personal genome sequenced. One is science writer Carl Zimmer, who jumped at the offer of having the company Illumina sequence his genes as part of an educational program. “It was like saying, ‘Would you like to go to Jupiter?’” he said. When he took his genome data to geneticists for analysis, he was treated to the usual hype. One scientist even pointed at him and said “That is not Carl Zimmer,” and then pointed to the hard drive containing his genome data and said, “This is Carl Zimmer.”
But when the data was analyzed, there was nothing much to see. Zimmer learnt he has a mutation that puts him at risk for high cholesterol, and that his genes give him powerful muscle fibers. It was all pretty underwhelming. Oh, but look! He has 2% Neanderthal genes! Surely that’s interesting? Not really—most people outside of Africa have similar levels of Neanderthal in them.
This boring profile, Zimmer knew, was actually good news. He didn’t have the genes for early-onset dementia, or a high risk of cancer, or diabetes, say. But if he did, it might not have been news to him, anyway. Such predispositions are generally passed along in a family’s genes, so they tend to be anticipated without genomic analysis. “Apart from a few well-known cases, most genetic information has only a very mild effect on the person,” says Herrero.
The limited information we receive reflects the fact that there’s still a lot we don’t know about what specific genes do, or which genes underlie which inheritable traits. What’s more, budget DNA tests deliver information on just a handful of known genetic linkages, such as some genes related to Alzheimer’s, particular cancer variants, or obesity. To have your full genome read will set you back a thousand dollars or more.
But another part of gene sequencing’s limitation is intrinsic to the way genes work. Many traits we might care about—such as height and intelligence, as well as many diseases with an inheritable component like heart disease and Type 2 diabetes—are thought to be influenced by a broad suite of genes. Not only does each gene have an almost negligible impact individually, but even collectively, their effects appear to be tiny and easily swamped by environmental factors such as lifestyle and upbringing. What’s more, such environmental influences can regulate how active a gene is.
Take genes linked to intelligence, for example (which are, whatever some researchers say, certainly not “intelligence genes”): These have so far only been able to account for a meager few percent of variations in IQ. Because genes interact with one another in complicated networks, the effects of individual genes on many traits are so highly dispersed that some researchers have even suggested it might be safer to make the default assumption that many diseases are ‘omnigenic’: that they involve the entire genome.
Having a gene like the cancer-linked BRCA1 mutation doesn’t at all condemn you to getting the affliction it’s associated with. What’s more, having a gene like the cancer-linked BRCA1 mutation doesn’t at all condemn you to getting the affliction it’s associated with. A few disease-linked gene variants, like the one that causes cystic fibrosis, will definitely lead to the disease, but most are just associated with increased statistical risk. This might be valuable information (because it might recommend a particular diet, say) but it is in no way a signature of your destiny. And if the mutation you carry is recessive, it may not ever present itself: Having one copy of the gene out of the two that every genome contains isn’t a problem, because the other healthy variant predominates. Indeed, it’s thought that perhaps one in five healthy people carry a rare disease mutation—but such headlines are not as alarming as they might sound.
Data interpretation? You’re on your own
As personal genome analysis becomes more common, the companies offering such services need to help people understand these issues rather than implying that they offer a “readout of you.”
Health counselors are struggling to give good advice to people coming to them with limited (but seemingly imperative) information about their genome. One cancer expert told me that an automatic mastectomy for carriers of the BRCA1 mutation—the option taken by Angelina Jolie—is not what he’d always recommend. But perception of risk is notorious for straining our intuition—just look at how people weigh up the risks of flying and driving.
Genome-sequencing companies don’t do much to help allay the hysteria. They typically deliver risk analyses directly to customers to make of what they will, not to their better-informed doctors. When British tech writer Martin Robbins had his genome analysed in 2010 by the California-based company Navigenics, his results included what Robbins dubbed a “wall of death”: a string of above-average genetic risks for Alzheimer’s, prostate cancer, and atrial fibrillation. All he was given along with the concerning results was prophylactic advice that could be charitably described as obvious: exercise, stay in shape, don’t smoke or drink too much.
After doing business for several years, 23and Me finally acceded to the US Food and Drugs Administration’s demands that they stop selling their unvalidated tests in their then-current form. Critics said that the company wasn’t applying any real critical judgement to the literature on gene-disease associations used for their risk analyses. If a link had been reported, they just accepted it for inclusion in their customer reports, even though many such studies were statistically weak and their claims have ultimately proved groundless. (The company denied that there was any need for regulation of such private genetic services.)
After doing business for several years, 23andMe acceded to the US Food and Drugs Administration’s demand that they stop selling their tests in their then-unregulated form. “We have never denied the need for regulation, and indeed were engaged with the FDA for more than two years before the 2013 letter,” said a 23andMe spokesperson. Critics said that the company wasn’t applying any real critical judgement to the literature on gene-disease associations used for their risk analyses. The accusation was that if a link had been reported, they just accepted it for inclusion in their customer reports, even though many such studies were statistically weak and their claims have ultimately proved groundless. The spokesperson denies this, saying that the company has “always applied rigorous standards for the studies we use to support our reports regarding study size, replication, and publication in reputable peer-reviewed journals.”
Commercialized personal DNA testing should prompt a proper (and overdue) discussion of these issues. Genomic data will be valuable for biomedical research, but it’s not likely to satisfy our craving to know ourselves. The more we appreciate that, the better we’ll be able to place genes in their true context.
“I’ve learned just how hard it remains for experts to make sense of anyone’s genome,” Zimmer wrote. Any “sense” you’ll make out of it will stop well short of a sense of you.
So go ahead and get your genome read—and learn from the disappointment.
Correction: This post has been updated to include a statement from 23andMe.