Page 44 - Forbes - USA (November 2019)
P. 44
Daphne Koller Cont. weeks instead of years,” Koller says.
N made in heaven” for investors, she says. Within six
AI plus biology, her background, was a “marriage
months Koller raised $100 million from ARCH
Ventures, Andreessen Horowitz, Foresite Capital,
40 Alphabet’s venture fund GV and Third Rock, with
Jeff Bezos and others joining later. In April, she
N HOW TO PLAY IT landed a deal with Gilead Sciences that gives Insi-
According to
O tro $15 million now with $1 billion to follow if it
I Gary Robinson helps find a treatment for a deadly form of nonal-
V A T Quantum com- coholic fatty liver disease. The disease is expected to
O puting and AI are soon become the leading cause of liver transplants.
N massive tailwinds
N Not many scientists get for healthcare “There are very few individuals who understand
I research and San
solicited for photo ops, but for Daphne Koller it’s a both sides of the beast,” says Mani Subramanian,
R Diego’s Illumina
O regular occurrence. “It happens at pretty much any leads the way. Its who heads liver disease clinical research at Gilead.
F
event that has tech people,” Koller says when asked machines have “The biology as well as the deep learning.”
N lowered the cost
O about one recent snapshot. “It’s a little awkward. of sequencing a Insitro’s future payouts from Gilead hang on
I whether it can identify five proteins that could be
T It’s not like I feel like this is something I deserve.” human genome
P
I Selfie requests are just one sign of Koller’s star- from $10 million targets for drugs and then whether targeting those
R in 2007 to $1,000
C dom, earned from more than 20 years bridging and are chang- proteins leads to approved therapies for the liver
S computer science, biology and education. She disease. The contingent payments, which include
E ing how cancer
R chalked up a string of accolades along the way: screening and revenue sharing from successful drugs, helped Insi-
P
getting a master’s degree from Jerusalem’s He- research is done. tro earn a spot on Forbes’ inaugural AI 50 list of the
— “We are moving
most promising artificial intelligence companies.
Y brew University at 18; becoming a Stanford Uni- from a world
G versity professor focused on machine learning at where decisions More than 20 other startups are chasing the
O on which drugs
L 26; winning, nearly a decade later, a Mac Arthur to give a patient dream of faster, cheaper drug discovery through
O were primarily AI. Among them are Notable Labs, with $55 mil-
N “genius grant” for research that combined artifi-
H cial intelligence and genomics; cofounding $1 bil- made on edu- lion of venture capital, and Verge Genomics, with
C cated guesses
E lion (valuation) Coursera, an early platform to let to one where $36 million. Novartis has announced a five-year
T
people around the world take university classes they are made AI collaboration with Microsoft, and Merck and
on the basis of
for free. data,” says Gary GSK have startup partnerships as well.
The next act for this 51-year-old innovator: In- Robinson, a port- Artificial intelligence does not make biology
folio manager
sitro, a firm in South San Francisco that aims at $260-billion- easy. “I don’t think the platform can be magic,”
to find new drugs by sorting through masses of in-assets Baillie Koller says.
data. If it succeeds, it will have overturned how Gifford. A recent Before Insitro can reap rewards, a few hundred
dip in sales
drugs get discovered. thousand lab tests need to happen. Koller has the
growth caused
Lab biologists typically focus on a few specific Illumina shares to energy. Bouncing around Insitro’s office—she
proteins as drug targets. If those fail, data scientists drop by 25% from gave away her desk chair to one of her 53 employ-
record highs, but
make suggestions for others to try. Insitro, on the Robinson shrugs ees because she never used it—she moves from
other hand, wants to collect much more data be- at the volatility. a room named Macrophage (a white blood cell)
fore the biologists go off on their hunt. It will le- “The healthcare to one named Elastic Net (a data-modeling tech-
sector is large
verage advances in bioengineering (such as Crispr and inefficient nique) to show off the latest lab equipment.
gene editing) and in software that enables comput- and therefore it is Big Pharma’s interest would seem to make In-
ripe for change,”
ers to see things that escape humans. sitro a likely acquisition target if it hits pay dirt.
he says, “Illumina
Koller describes her aha moment this way: is the primary But Koller says she doesn’t want to see Insitro
“Machine learning is now doing amazing things beneficiary.” “swallowed into the maw” of a larger organiza-
if you give it enough data. We finally have the op- tion. She wants it to make its own branded drugs.
portunity to create biological data at scale.” The ultimate goal is that the people asking for HOW TO PLAY IT WRITTEN BY ANTOINE GARA; PATRICK WELSH FOR FORBES
Insitro’s computational experts and biologists photos ops will be healthier thanks to Insitro.
work together to create lab experiments to pro- Koller says she hopes they come up to her and
duce massive custom data sets. Machine learning say, “Because of you, I have my life back.” F
models then find patterns to suggest new tests
and potential therapies. Robotics like automat- F IN AL T HO UG HT
ed pipetting machines reduce human error. With “DATA TRUMPS EVERYTHING.”
all this, Insitro can do “experiments in a matter of —Josh Estelle, a lead engineer for Google Translate
F O R B E S . C O M N O V E M B E R 3 0 , 2 0 1 9

