Q&A: Former Genentech Exec Scheller Discusses His Plans for 23andMe's New Therapeutics Division

NEW YORK (GenomeWeb) – Drug development is an expensive business fraught with failure, but consumer genomics firm 23andMe wants in.

The Tufts Center for the Study of Drug Development last year estimated that the cost of getting a new drug approved – after factoring in failures and capital costs – is $2.6 billion. In 2003, the same group had pegged the cost per new approved drug at $800 million. 

So, it's not a surprise that 23andMe has hired former Genentech executive Richard Scheller, no stranger to the risks involved in drug development, to lead its new therapeutics division. Last year, he retired from Genentech, where he led basic research in oncology, immunology, neuroscience, and infectious disease.

Most importantly, Scheller believes, as 23andMe does, that genomics will transform medicine. He received the Albert Lasker Basic Medical Research Award for his work identifying the molecular and regulatory mechanisms involved in the rapid release of neurotransmitters. And for 19 years at Stanford University he taught in the department of biological sciences and the department of molecular and cellular physiology.

Scheller discussed with GenomeWeb why he joined 23andMe after leaving Genentech and his plans for the new therapeutics division. Below is a transcript of the interview, which has been edited for clarity.

Why did you decide to join 23andMe from Genentech? Had you worked with 23andMe before?

I've known about 23andMe since their founding. I was head of research seven or eight years ago when Genentech made an investment in 23andMe through a Series A funding. I have followed the company ever since, and one of the last deals that I signed off on as head of research and early development at Genentech was a collaboration on Parkinson's disease. So, I believe that human genetics will play an ever increasing role in the identification and development of therapeutics. As you know, 23andMe has the world's largest human genetics database. So, it seemed like an interesting new challenge to try and put the two together.

Regarding the database [containing 850,000 genotyped customers and a trove of phenotypic data collected through questionnaires], a large proportion of 23andMe's customers have given consent to have their data used for research. Do you think you'll need to issue new consent procedures to use customers' information for internal drug development efforts?

In terms of mining the data to try and identify drug targets I don't believe we will need new consent. If we were to somehow engage a person in a clinical experiment, which of course would be several years down the line, that would take consent.

Beyond the database, what other resources and tools are you going to bring into the company to build out the therapeutics division? Do you have any molecules, for example, to start off with? What other technologies will you invest in?

The first thing to do is to work with the database and try and identify targets. Then, we will furnish out lab space and [put in] all of the technologies that are used for modern drug development. It's a daunting task but one that I look forward to.

It is, as you say, a daunting task and an expensive task. How is this being funded?

Well, 23andMe is well funded and we have a plan to be able to fund the next several years of work in the therapeutics unit. … I don't even start at 23andMe until April 1. So, it isn't quite clear to me exactly what we're publicly discussing in terms of financing and so on.

In announcing the effort, 23andMe said it is planning to advance drugs for rare and common diseases. Are there any particular disease areas you're focusing on as a start?

We plan on being opportunistic there and we'll be looking at immunological disorders, metabolic disorders, neuroscience, infectious disease, oncology, ocular disease, and so on. We don't plan on restricting ourselves early on. We want to let the database lead the way.

Your first focus is to identify drug leads from the genomic data, but of course there are many other types of information that would go into developing personalized treatments – assuming that's the goal here. Does 23andMe's database contain the diversity of data you'd need for this task and to take a systems approach?

We plan on combining the data from human genetics [in] the 23andMe database with the wealth of knowledge around the functions of various genes that exist, and it will be a combination of those approaches that will allow us to focus on particular targets.

23andMe has been inking collaborations with other pharma companies [such as Genentech and Pfizer]. Will this move to launch its own therapeutics unit be a positive for inking future collaborations with pharma or could some companies see this as a conflict of interest?

The collaborations that we have ongoing will be firewalled from the work that I'll be doing internally. This kind of thing happens all the time. Companies with databases or companies with a particular technology often work with multiple partners and have internal programs. So, that's not unusual. We're assessing how many new collaborations we will enter into, and we've talked to many of the partners and they're actually enthused. They know that as a small company, 23andMe won't be able to take projects into large clinical trials that are required for approval. It's extremely likely that some time in the future, we'll be looking for partners with the large pharmaceutical companies. So, they are quite excited about the prospect of partnering with us down the line if we have exciting therapeutics to bring forward.

The regulatory aspect of drug development is one that requires a lot of expertise and investment. 23andMe has seen its share of regulatory trouble [with its direct-to-consumer testing business], which the firm is still working through. Drug development being as highly regulated as it is, is this the right time for 23andMe to launch a therapeutics unit?

I've worked in that area for 14 years. So, I'm not naïve about the regulatory requirements, and I'm quite familiar with what's necessary to move ahead there. It will be some time before we need to interact with the FDA. That would be when we have a molecule we're excited about potentially bringing into man. And that will be some time in the future. So, within the therapeutics group, regulatory requirements are really not an issue in these early stages of research.

So, how big do you think the therapeutics unit will be? How many people do you plan to hire?

We haven't decided yet. We think the most important thing to do is to be, as I said, opportunistic. If we find a number of important targets … we will find the resources to invest in those targets. Until we've had time to really think through that and look through the database, the plans are really in flux.

Given your long experience in drug development, I'm sure you've seen your share of successes and failures. Many ambitious startups haven't succeeded in this business. What makes you think 23andMe will?

The major cost of drug development are the failures, and I believe that by going through a human genetics approach, one is likely to end up working on targets that are much more likely to succeed. So, it's the human genetics as the foundation for target discovery that gives me hope that we will be successful where others have failed.

A fire at a Manchester hospital may have destroyed lab equipment and data, the Guardian reports.

Researchers generate a genetic database from skeletal remains from the 1845 Franklin Expedition to the Arctic, Live Science reports.

Researchers in China have begun another trial using CRISPR/Cas9 approaches in cancer patients, according to the Wall Street Journal.

In Science this week: human DNA found in sediments from archeological sites lacking bones, and more.