Researchers at the genetics company 23andMe have identified 15 regions of the human genome that influence whether someone considers themselves a "morning person," meaning someone who likes to wake up and go to bed early, or an "evening person." The finding, published today in Nature Communications, means that there might actually be a genetic basis for certain people's tendency to wake up at the crack of dawn — or to stay in bed past noon.

Until recently, most of the work done on the genetics of being an early riser was conducted in fruit flies or mice. Last year, for instance, scientists found about 80 genes that can influence whether a fly prefers to be active in the morning or later in the day. But exactly how that information translates to a large group of humans wasn't entirely clear. Now, 23andMe geneticists say that data from 90,000 of their customers has helped them begin to figure that out. The study isn't perfect, but if researchers can push the findings further, that information could one day have some medical benefits.

"We think of our preferences as things that we come up with — things that are kind of spontaneous parts of who we are — but they do have a basis in biology," says David Hinds, a statistical geneticist at 23andMe and a co-author of the study. "I think it's just very interesting for people to see how their biology influences who they are."

To figure out which regions of the human genome are associated with being a morning person or an evening person, the researchers first asked consenting 23andMe customers to answer two questions about being a morning or an evening person. Then, they analyzed each customer's answers in relation to their genome. "We have data for 15 million genetic variants, and we just compare each of those variants and say, 'Is this variant predictive of how those participants answered that question?'" Hinds says.

Hinds and his team found that the variants that could be used to predict people's preferences were clustered in 15 regions of the genome. When the scientists looked at the individual effect of these genetic variants, they found that each had a relatively small effect. "Each of these might shift your chances of being a morning person by between 5 and 25 percent," Hinds says. But when they combined the effect from all the regions, Hinds and his team found that the respondents who had variants that were most strongly linked with being a morning person were about twice as likely to consider themselves early risers as the people with the lowest scores.

The researchers also looked at what roles these regions play in the human body. They found that seven of the 15 regions were close to genes known to be involved in our circadian rhythms — the physical, behavioral, and mental changes that align roughly with a 24-hour cycle. The researchers also found that some of the regions they identified are involved in phototransduction, or the process by which our eyes convert light into a signal that can go to the brain.

"The associations seem to be very strong," says Eran Tauber, a geneticist at the University of Leicester who didn't work on the 23andMe study, but who was involved in last year's fruit fly experiment. "And the fact that some of these genes were already implicated in the function of the circadian clock is very reassuring — although the study should be followed by more experimental work."

The study has a number of limitations. Because the scientists relied on an online survey, it's always possible that some people answered incorrectly. People don't always have the same definitions for the word "morning," says co-author Youna Hu, a data scientist who recently moved from 23andMe to Amazon. And others may have accidentally clicked on the wrong answer. To try to avoid confusion, the researchers didn't use data from people whose answers were incoherent from one question to the next, but some people may have slipped through the cracks.

The study also didn't take into account people's geographic location or the season during which respondents answered the questions. Given that the environment can have a very strong influence on a person's disease risk or traits — some researchers estimate that these are 50 percent influenced by environment — it would certainly help to have that information in a follow-up study. Finally, the researchers also weren't able to pinpoint the exact genes that underly the association, but they're hopeful that they might be able to get there soon.

Now that the study has been published, Hinds and his team would like to combine their data set with one from a British research group that has been working on similar questions using the UK Biobank. They think that kind of collaboration could yield more robust results. Already, the 23andMe researchers have found that the "data is quite consistent" with theirs, Hinds says.

"The take-away message is that there is a significant genetic basis for morning and evening preference," Tauber says. Armed with that information, scientists might one day be able to determine the best time to administer medical treatments like chemotherapy, he says. That's because finding the "optimal time would allow the use of lower doses and would minimize side effects." And of course, identifying the genes that are associated with these preferences could also help scientists understand the basic biology that underlies what humans think of themselves. "The potential for this knowledge is huge," Tauber says.

Verge Video: Debunking the "mad genius" myth