Venter gains a key human asset

 

“One of the biggest challenges in medicine and science today—and thus, one of HLI’s biggest challenges—is how to interpret the vast amount of biological data we are generating from sequencing individual genomes,” said Venter in a news release. “To make these data interpretable and clinically actionable will require new computational tools. Franz brings not only unquestionable talent in this area, but also a fresh perspective and a creative mind to tackle what has never before been attempted. We are pleased to have Franz join us in our quest to revolutionize healthcare,” he concluded.

 

HLI is building the world’s most comprehensive human genotype and phenotype database to tackle the diseases associated with aging-related human biological decline. Building on Och’s expertise in language translation and using his skills in machine learning, he will be responsible for developing new computational methods to translate the human biological information in the forms of the whole human genome, microbiome and protein sequencing data into the language of human health and disease using human phenotype data.

 

Using language translation as a stepping-off point for HLI’s incredibly ambitious genomics project is not as far-fetched as it may sound at first, according to Dr. Clifford Baron, vice president and chief operating officer at CollabRx, a San Francisco-based data analytics company that uses cloud-based expert systems to inform healthcare decision-making by aggregating and contextualizing the world’s knowledge on molecular medicine with specific insights from the nation’s top clinical experts.

 

Baron points out that language, like the human genome, is “highly contextual. What is ‘right’ is dependent on what’s around; in the case of the genome, codons, transcripts and proteins, for example.” Baron is familiar with Venter’s background, capabilities and financial resourcefulness. “The field needs another moon shot,” Baron adds, and in that regard he calls Och “an intriguing hire.”

 

“Applying machine learning to genomics is not a novel concept,” notes Dr. Boris Umylny, CEO at Lumenogix Inc., a company that provides bioinformatics for biologists.  “It has been used by ENCODE for years. Recently, IBM and the New York Genome Center announced that they will test a prototype of Watson developed specifically to address genomic issues. Longevity also is not new; Methuselah Foundation has been at it for more than a decade.” But, Umylny adds, “It is novel to see such prominent startups in the field.

 

“In my opinion, Dr. Venter's initiative is more technically sound than Google's Calico. With up to five HiSeq X Ten [genome sequencers], they will be able to get a much better look at the human genome than Google with 23AndMe data.”

 

Ramon Felciano, a founder of Ingenuity Systems (acquired by QIAGEN in 2013), where he is now vice president of strategy and technology for bioinformatics, opines that there will be “a lot of value out of big data, but the clinical context will always be important, with each patient’s phenotype revealing a constellation of information,” for the interpretation and analysis of complex biological systems. For clinical applications, he says, understanding how new data compares to what is currently known will be key. “It’s great to see Venter tackling this,” he adds.

 

Och is currently recruiting and building a team of research scientists and software engineers who will be located in Mountain View, Calif.

 

“We’re going to need the best and brightest from the areas of computer science, machine learning and big-data generation and interpretation, as well as those from biology, genomics and bioinformatics, to reach a new level of understanding of this massive database,” said Och. “I look forward to working with Craig and the team at HLI to enhance our understanding of human biology, to better manage the healthy aging process and thus increase the healthy human lifespan.”

 

Och will be working closely with Dr. Yaron Turpaz, HLI’s chief information officer, who leads bioinformatics efforts at HLI. The new Mountain View facility headed by Och will complement the informatics program at HLI’s La Jolla-based facility in San Diego and the computing and informatics program and facility in Singapore headed by Turpaz.

 

Since 2004, Och has been a research scientist and manager leading Google’s Machine Translation Group. Och, who has been working on statistical machine translation since 1997, joined Google to research new ways by which Google could offer a much better machine translation system to their users by employing an approach that uses massive amount of data. In just a few years he and his team have revolutionized online translation. Now more than 80 different languages, including Bengali, Basque, Swahili, Yiddish and Esperanto, can be translated. Today there are more than 200 million active users on Google Translate, and Och estimates that most of the translation done on the planet is done through Google’s system.

 

Before joining Google, Och was a researcher in the Information Sciences Institute at the University of Southern California, working on Defense Advanced Research Projects Agency projects related to language translation. He holds a master of science degree in computer science from FAU Erlangen-Nuremberg in Germany and a Ph.D. in computer science from RWTH in Aachen, Germany.

 

HLI, a privately held company headquartered in San Diego, was founded in 2013. Using advances in genomic sequencing, the human microbiome, proteomics, informatics, computing and cell-therapy technologies, HLI seeks to build a comprehensive database of human genotypes and phenotypes as a basis for a variety of commercialization opportunities to help solve aging-related disease and human biological decline. HLI will be licensing access to its database and developing new diagnostics and therapeutics as part of its product offerings.