This story is over 5 years old
Tech by VICE

Google Translate's Lead Scientist Is Going to Help Conquer Death

After all, translating genetic code into actionable information isn't all that different from translating languages.

by Meghan Neal
Jul 30 2014, 8:10pm

Image: Shutterstock

Google, it turns out, doesn't have a monopoly on the ambitious goal to conquer death. In fact, the tech giant just lost one of its top computer scientists to a genetics startup called Human Longevity, which is working to achieve exactly what the name implies.

It's the latest project being spearheaded by Craig Venter, the biologist famous for being one of the first to sequence the entire human genome. Now, armed with advanced technology that brought the price tag and time suck of genome sequencing way down, Venter's team will sequence tens of thousands of genomes every year.

The firm will also sequence the genomes of the trillions of microbes that live on and inside the body—bacteria, viruses, and fungi—and integrate that into the data. The hope is that the resulting torrent of genetic information could hold the key to preventing disease and slowing aging—if someone can figure out how to decipher the massive database. Or more accurately, teach a machine to do it.

And that's why Human Longevity announced yesterday it's hired the former head of Google Translate, Franz Och, to develop new computational tools and algorithms to translate the bio data into actionable insights—like developing personalized medicine, early diagnosis, cures for killers like cancer and heart disease, and eventually stem cell therapies. 

In a way, interpreting the jumbled DNA letters that make up our genetic code is a lot like translating a language. Each sequenced genome is made of some 3 billion As, Cs, Gs, and Ts strung together.

"Imagine the genome as a book written without capitalization or punctuation, without breaks between words, sentences, or paragraphs, and with strings of nonsense letters scattered between and even within sentence," Genome News Network explains. "Even in a familiar language it is difficult to pick out the meaning of the passage."

Indeed, it would be impossible for humans to crack the code; it'll take an intelligent machine to analyze the data and unlock the biological secrets hidden inside. "Your genetic code translates into your biological code which translates into you," Venter told U-T San Diego, where the startup is based. "We need to use machine learning to find associations between genes that mere mortals can't find from staring at the data. It's too complex."

Of course, analyzing big data is Google's bread and butter. Google Translate can instantaneously translate text to 80 different languages using data-based machine learning, with relative accuracy. The machine is fed hundreds of thousands of documents previously translated by humans and learns to recognize patterns to apply to future translations. The more data it's fed, the better it gets.

Meanwhile, while the genetics company is busy writing algorithms, Google is gearing up to do some genome sequencing of its own. Venter's startup sprung up just seven months after Google Calico launched, a venture project that's also setting out to fight aging with technology.

Then, last week, Google announced its Baseline Study, which will collect the genetic information from 175 people (later thousands) to create a comprehensive picture of what a normal, healthy human should look like. The idea is, once you have a clean slate it's easy to pinpoint disease biomarkers as health deteriorates.

Google staffed up with biologists to lead the projects and will sequence anonymous genomes for the purpose of the research. In the future, it could also leverage the onslaught of data streaming in from wearable fitness gadgets and health apps. When announcing the Baseline Study, the company offered assurances that the data wouldn't be used for commercial purposes and would stay anonymous, but the project is raising privacy flags anyway, and for good reason.

Venter, for instance, made no such promises, and in fact is open to selling the genetic information and insights to, say, pharmaceutical companies. He's poured some $30 million in startup capital into the company already (the fancy Illumina machines that make it possible to sequence a genome super-quick cost a million each; Human Longevity bought 20).

But the money-making potential of the elusive fountain of youth is just about impossible to overstate. The race to kill death is on.