A Potential Problem For The Business Of Genomic Prediction

Hopefully, Gwyneth Paltrow won’t get involved in this (boldface mine):

Recently, Vitaliy Husar received results from a DNA screening that changed his life. It wasn’t a gene that suggested a high likelihood of cancer or a shocking revelation about his family tree. It was his diet. It was all wrong.

That was, at least, according to DNA Lifestyle Coach, a startup that offers consumers advice on diet, exercise and other aspects of daily life based on genetics alone. Husar, a 38-year-old telecom salesman, had spent most of his life eating the sort of Eastern European fare typical of his native Ukraine: lots of meat, potatoes, salt and saturated fats. DNA Lifestyle Coach suggested his body might appreciate a more Mediterranean diet instead.

“They show you which genes are linked to what traits, and link you to the research,” Husar told Gizmodo. “There is science behind it.”

DNA Lifestyle Coach isn’t the only company hoping to turn our genetics into a lifestyle product. In the past decade, DNA sequencing has gotten really, really cheap, positioning genetics to become the next big consumer health craze. The sales pitch—a roadmap for life encoded in your very own DNA—can be hard to resist. But scientists are skeptical that we’ve decrypted enough about the human genome to turn strings of As, Ts, Cs and Gs into useful personalized lifestyle advice.

Indeed, that lifestyle advice has a tendency to sound more like it was divined from a health-conscious oracle than from actual science. Take, for instance, DNA Lifestyle Coach’s recommendation that one client “drink 750ml of cloudy apple juice everyday to lose body fat.”

“Millions of people have had genotyping done, but few people have had their whole genome sequenced,” Eric Topol, a geneticist at Scripps in San Diego, told Gizmodo. Most consumer DNA testing companies, like 23andMe, offer genotyping, which examines small snippets of DNA for well-studied variations. Genome sequencing, on the other hand, decodes a person’s entire genetic makeup. In many cases, there just isn’t enough science concerning the genes in question to accurately predict, say, whether you should steer clear of carbs.

We need billions of people to get their genome sequenced to be able to give people information like what kind of diet to follow,” Topol said.

Here’s what I don’t get: any genomic information these companies–note the word companies–acquire will be proprietary. They’re not going to share this information with each other. So unless there’s a ‘big dog’ that has the lion’s share of genomes (we do like mixing our metaphors), how do these companies expect to acquire enough genomes to say anything meaningful? And they won’t have collected the same patient metadata in the same way either?

Though maybe GOOP! should get into this business:

When asked whether it was possible that DNA Lifestyle Coach’s claims might have any validity, Topol laughed.

One day, he said, it’s likely we’ll have some genomic insight into what types of diets are better suited for certain people. But, he added, it’s unlikely that we will ever accurately predict the sort of granular details DNA Lifestyle Coach hopes to, like exactly what SPF of sunscreen you should be using on your skin.

“There are limits,” he said.

This entry was posted in Bidness, Genomics. Bookmark the permalink.