Member News

March 03, 2017

Joe Betts-Lacroix and Tim Robertson want to build a better mouse trap — so to speak — and in the process revolutionize the way laboratories collect data from animals.

But the duo's "digital vivarium" company, called Vium Inc., is part of a larger movement by traditional tech players into the life sciences, with many of those companies led by Bay Area venture capital firm Data Collective.

The idea isn't simply to gather more data to reshape the speed and cost of drug development. Instead, Data Collective's portfolio of companies are largely focused on using machine learning, artificial intelligence and algorithms to gather better data and make better decisions on, for example, moving forward with a drug.

"We believe years can be taken out of the process, if only scientists had the right technology to do so," said Armen Vidian, an operating partner with Data Collective. "Scientists should be equipped with the same flexibility and power than any IT professional has."

Data Collective, which has offices in Palo Alto and San Francisco, over the past two years has been the most active of traditionally tech-focused VCs in the world of medical diagnostics and tools. Its bets are relatively small — they generally aren't supporting massive drug trials — but could have big payoffs as the the tech and biotech worlds collide.

A recent report from Silicon Valley Bank noted that Data Collective made eight diagnostics/tools investments in 2015-16, followed by five by traditional tech investor Khosla Ventures.

"The pace of advancement and the lack of regulatory hurdles makes advancement faster," said Jonathan Norris, a managing partner at Silicon Valley Bank's healthcare division.

Data Collective's investments range from San Mateo's Vium, which is working on sensors that gather around-the-clock data from lab animals, and 3Scan Inc., a San Francisco company developing technology to section and image tissue in pathology labs, to Emeryville next-generation DNA testing company Diassess and Atomwise, a San Francisco company applying artificial intelligence to drug discovery.

Those types of companies, Vidian said, could play a role in one of the drug development world's biggest pushes: personalized medicine. The key to getting the right drug to the right person at the right time is collecting deep data that pass muster with scientists, drug developers and regulators.

But tech investors and companies — such as South San Francisco's Verily Life Sciences, spun out of Alphabet/Google (NASDAQ: GOOG) — have been criticized for a general philosophy of breaking things and failing quickly to get to answers quickly. Making various versions of software is vastly different than working through the scientifically rigorous and highly regulated process of drug development.

"There clearly is room for fresh thinking and fresh ideas around innovation, and there's a lot to be learned from the success of the tech industry," said Matthew Hudes, the former managing partner of consulting firm Deloitte's biotechnology practice who now leads bdlBiologix. "But we shouldn't forget there have been large failures over there as well.

"We all are learning we ought to integrate."

There is no denying that traditionally tech-focused VCs and companies — from Google Ventures and General Electric to AME Cloud Ventures and Andreessen Horowitz — have taken a shining to biotech and its decades of laborious procedures.

Five-year-old Data Collective, which late last year closed its fourth fund at $212 million for Big Data and IT infrastructure seed and early-stage, is clearly focused on traditional tech ventures. The firm's life sciences investments — more than a dozen in the Bay Area — build on the expertise collected by managing partners Matthew Ocko and Zachary Bogue.

Bogue, an angel investor in Uber and Square, focuses on space informatics, machine learning in the financial services and risk management systems with an artificial intelligence backbone; Ocko's interests run from synthetic biology to robotics.

The firm's life sciences investments have a high bar, Vidian said: Big Pharma companies need strong, reproducible data to justify moving projects forward into human clinical trials, not video games.

"The bar is high because having confidence in the data is important so that the most effective compounds make it through development — and the ineffective ones are screened out quickly," he said. "So getting highly reproducible results consistently is critical."

"Coming into this with no understanding of biology is tough," said Vidian, who graduated M.I.T. with a degree in mechanical engineering, Stanford with a master's in management science and worked for a series of medical device companies. "We look for teams with strong computing as well as biology experience."

Take Vium, which in June 2016 received $33 million from Data Collective, Lux Capital, Founders Fund, Dolby Family Ventures and AME Cloud Ventures.

When researchers reach into a mouse cage to check on heart rate and respiration or other metrics needed from an experiment, the animals' hearts beat more quickly and their breathing changes. Sometimes whether the researcher is male or female makes a difference.

That is a problem that adds noise to lab data, said Betts-Lacroix, Vium's chief technology officer, and Robertson, the CEO. They believe they can minimize the interactions between human and animal by using sensors and processors that collect "cleaner" data, then pushing the information into the cloud and running it through algorithms.

Sensors, video trackers and processors can give a better read, for example, that a mouse is drowsy from a drug when the animal otherwise appears fine.

"It's all about the lab and clinical results and translatability," Robertson said.

Four of the 10 largest pharmaceutical companies are using the system from 55-employee Vium, he said. The interest for those companies is in squeezing development costs: By some calculations, every day a blockbuster drug isn't on the market costs a company about $30 million a day.

"Any way that tech helps humans is good," Betts-Lacroix said.

Indeed, Silicon Valley Bank's Norris said, tech investors such as Data Collective are bringing a fresh eye to life sciences problems.

"It's fascinating to see where the tech investors see all the advancement in collecting data and applying it in health," Norris said. "Where you get a lot of data from sequencing and drill down and figure out the biomarkers and the type of technology applied there is amazing."

DATA COLLECTIVE'S PORTFOLIO

Among Data Collective's Bay Area life sciences investments are:

3Scan Inc.(San Francisco) — developing technology to section and image tissue in pathology labs

Agenovir Corp.(South San Francisco) — gene editing to create topical treatments for immune-system weakening viruses

Atomwise(San Francisco) — deep-learning technology for quick small-molecule discovery

CloudMedx Inc.(Palo Alto) — clinical algorithms, machine learning, advanced natural language processing to help doctors develop care plans for patients

Cofactor Genomics Inc.(San Francisco) — tapping machine learning technology to sample RNA to diagnose disease

Diassess(Emeryville) — next-generation DNA testing with applications across health care and agriculture

Enlitic Inc.(San Francisco) — deep-learning networks examine millions of images to automatically learn how to recognize diseases

Freenome Inc.(South San Francisco) — liquid biopsy company for early detection of cancer

Pivot Bio Inc.(Emeryville) — microbes that can help crops grow better, protect against pests and improve crop quality

Transcriptic Inc.(Menlo Park) — a robotic cloud lab to advance drug discovery and synthetic biology

Vium Inc.(San Mateo) — using sensors and processors to create a "digital vivarium" for deeper, more meaningful data from animals

Whole Biome Inc.(San Francisco) — analyzing the microbiome, the community of small organisms living in or on our bodies

Zymergen Inc.(Emeryville) — engineering better microbes for useful molecules that have better predictability