THE NEW YORK TIMES, TUESDAY, DECEMBER 1, 2020 N D3Mars was once home to seas and oceans,
and perhaps even life. But our neighboring
world has long since dried up and its atmos-
phere has been blown away, while most ac-
tivity beneath its surface has long ceased.
It’s a dead planet.
Or is it?
Previous research has hinted at volcanic
eruptions on Mars 2.5 million years ago. But
a new paper suggests an eruption occurred
as recently as 53,000 years ago in a region
called Cerberus Fossae, which would be the
youngest known volcanic eruption on Mars.
That drives home the prospect that beneath
its rusty surface pocked with gigantic volca-
noes that have gone silent, some volcanism
still erupts to the surface at rare intervals.
“If this deposit is of volcanic origin then
the Cerberus Fossae region may not be ex-
tinct and Mars may still be volcanically ac-
tive today,” scientists at the University of
Arizona and the Smithsonian Institution,
write in their paper — which was posted on-
line ahead of peer review and has been sub-
mitted to the journal Icarus.
The site of the potential eruption, seen in
images from Martian orbit, is near a large
volcano called Elysium Mons. It is about
1,000 miles east of NASA’s stationary In-
Sight lander, which touched down on Mars
in 2018 to study tectonic activity on the red
planet. Appearing like a crack in the sur-
face, the feature looks like a recent fissure
eruption, where subsurface volcanic activi-
ty has caused superheated volcanic ash and
dust to burst through the surface. It is simi-
lar to deposits caused by pyroclastic erup-
tions that scientists have spotted on the
moon, Mercury and Earth.
Originating from magma deep beneath
the surface, the eruption would have
reached a height of several miles before fall-
ing back to the ground. The amount of ma-
terial is estimated at 100 times less than the
eruption of Mount St. Helens in 1980, said
Steven Anderson, an earth sciences profes-
sor at the University of Northern Colorado
in Greeley, who was not involved in the pa-
per.
It is the presence of darker material here,
coupled with its symmetrical appearance
around the fissure, that hints at an eruption.
Known as a fault scarp, this type of feature
is “very common in Hawaii” as magma near
volcanoes causes the surface to expand and
crack, said Robert Craddock from the
Smithsonian Institution, a co-author on the
paper.
By counting the number of craters visible
around the feature and in the deposit itself,
which is roughly six miles across, the team
date the potential eruption ranging from
53,000 to 210,000 years ago. This would by
far be the youngest known volcanic erup-
tion on Mars.
“I think it’s pretty compelling,” Dr. Ander-
son said.
If it holds up to scrutiny, the discovery
would have large implications for Mars. In
geological terms, 53,000 years is the blink of
an eye, suggesting Mars might well still be
volcanically active now. It could also havebig implications for the search for life on
Mars.
Such volcanic activity could melt subsur-
face ice, providing a potential habitable en-
vironment for living things.
“To have life, you need energy, carbon,
water and nutrients,” Dr. Anderson said.
“And a volcanic system provides all of
those.”
NASA’s InSight lander may have already
recorded activity linked to this site. Using a
seismometer, it has measured hundreds of
“marsquakes” or vibrations in the Martian
surface. But only two of these have been lo-
calized — and both came from Cerberus
Fossae.“It is certainly plausible that the tectonic
activity is related to volcanic activity,” said
Suzanne Smrekar from NASA’s Jet Propul-
sion Laboratory, who is the deputy principal
investigator on the InSight mission.
It might be possible for InSight to look for
more such activity soon.
“It is an exciting paper,” Dr. Smrekar said.
“Understanding the present day activity on
Mars is indeed a mystery and key to investi-
gating its evolution and habitability.”
Questions still remain, however. Lu Pan,
a planetary scientist from the University of
Copenhagen in Denmark, isn’t so sure
about the team’s dating method.
“If you want to date a very recent surface,
you rely on the population of small impact
craters,” Dr. Pan said. “And we have yet to
build this large database of small impact
craters.”
Even in a conservative scenario, howev-
er, David Horvath of the University of Ari-
zona, the paper’s lead author, said the erup-
tion would have been only a million years
ago. That alone would breathe new life into
our understanding of Mars.
“It definitely leaves open the possibility
that, deep in the surface, it may be active
today,” he said.Old Volcanic Activity on Mars Hints at a Habitat for Life
A new paper raises questions
about what may be on and
under the planet’s surface.
Top, an infrared image of the Elysium Mons volcano on Mars. Above, a deposit in
the Cerberus Fossae region that might be evidence of a volcanic eruption.NASA/JPL-CALTECH/ARIZONA STATE UNIVERSITY‘To have life, you need
energy, carbon, water
and nutrients.’
STEVEN ANDERSON
UNIVERSITY OF NORTHERN
COLORADOBy JONATHAN O’CALLAGHANNASA/JPL/MSSS/THE MURRAY LABSome scientists spend their lives trying to
pinpoint the shape of tiny proteins in the hu-
man body.
Proteins are the microscopic mecha-
nisms that drive the behavior of viruses,
bacteria, the human body and all living
things. They begin as strings of chemical
compounds, before twisting and folding into
three-dimensional shapes that define what
they can do — and what they cannot.
For biologists, identifying the precise
shape of a protein often requires months,
years or even decades of experimentation.
It requires skill, intelligence and more than
a little elbow grease. Sometimes they never
succeed.
Now, an artificial intelligence lab in Lon-
don has built a computer system that can do
the job in a few hours — perhaps even a few
minutes.
DeepMind, a lab owned by the same par-
ent company as Google, said on Monday
that its system, called AlphaFold, had
solved what is known as “the protein folding
problem.” Given the string of amino acids
that make up a protein, the system can rap-
idly and reliably predict its three-dimen-
sional shape.
This long-sought breakthrough could ac-
celerate the ability to understand diseases,
develop new medicines and unlock myster-
ies of the human body.
Computer scientists have struggled to
build such a system for more than 50 years.
For the past 25, they have measured and
compared their efforts through a global
competition called the Critical Assessment
of Structure Prediction, or C.A.S.P. Until
now, no contestant had even come close to
solving the problem.
DeepMind solved the problem with a
wide range of proteins, reaching an accura-
cy level that rivaled physical experiments.
Many scientists had assumed that moment
was still years, if not decades, away.
“I always hoped I would live to see this
day,” said John Moult, a professor at the
University of Maryland who helped create
C.A.S.P. in 1994 and continues to oversee the
biennial contest. “But it wasn’t always obvi-
ous I was going to make it.”
As part of this year’s C.A.S.P., Deep-
Mind’s technology was reviewed by Dr.
Moult and other researchers who oversee
the contest.
If DeepMind’s methods can be refined, he
and other researchers said, they could
speed the development of new drugs as well
as efforts to apply existing medications to
new viruses and diseases.
The breakthrough arrives too late to
make a significant impact on the coronavi-
rus. But researchers believe DeepMind’s
methods could accelerate the response to
future pandemics. Some believe it could
also help scientists gain a better under-
standing of genetic diseases along the lines
of Alzheimer’s or cystic fibrosis.
Still, experts cautioned that this technol-
ogy would affect only a small part of the
long process by which scientists identify
new medicines and analyze disease. It was
also unclear when or how DeepMind would
share its technology with other re-
searchers.
DeepMind is one of the key players in a
sweeping change that has spread across ac-
ademia, the tech industry and the medical
community over the past 10 years. Thanks
to an artificial intelligence technology
called a neural network, machines can now
learn to perform many tasks that were once
beyond their reach — and sometimes be-
yond the reach of humans.
A neural network is a mathematical sys-
tem loosely modeled on the network of neu-
rons in the human brain. It learns skills by
analyzing vast amounts of data. By pin-
pointing patterns in thousands of cat pho-
tos, for instance, it can learn to recognize a
cat.
This is the technology that recognizes
faces in the photos you post to Facebook,
identifies the commands you bark into your
smartphone and translates one language
into another on Skype and other services.
DeepMind is using this technology to pre-
dict the shape of proteins.
If scientists can predict the shape of a
protein in the human body, they can deter-
mine how other molecules will bind or phys-
ically attach to it. This is one way drugs are
developed: A drug binds to particular pro-
teins in your body and alters their behavior.
By analyzing thousands of known pro-
teins and their physical shapes, a neural
network can learn to predict the shapes of
others. In 2018, using this method, Deep-
Mind entered the C.A.S.P. contest for the
first time and its system outperformed all
other competitors, signaling a significant
shift. But its team of biologists, physicists
and computer scientists, led by a re-
searcher named John Jumper, were no-
where close to solving the ultimate prob-
lem.
In the two years since, Dr. Jumper and his
team designed an entirely new kind of neu-
ral network specifically for protein folding,
and this drove an enormous leap in accura-
cy. Their latest version provides a powerful,
if imperfect, solution to the protein folding
problem, said the DeepMind research sci-
entist Kathryn Tunyasuvunakool.
The system can accurately predict the
shape of a protein about two-thirds of the
time, according to the results of the C.A.S.P.
contest. And its mistakes with these pro-
teins are smaller than the width of an atom
— an error rate that rivals physical experi-
ments.
“Most atoms are within an atom diameter
of where they are in the experimental struc-
ture,” said Dr. Moult, the contest organizer.
“And with those that aren’t, there are other
possible explanations of the differences.”
Andrei Lupas, director of the department
of protein evolution at the Max Planck Insti-
tute for Developmental Biology in Ger-
many, is among those who worked with Al-
phaFold. He is part of a team that spent a
decade trying to determine the physical
shape of a particular protein in a tiny bacte-
rialike organism called an archaeon.
This protein straddles the membrane of
individual cells — part is inside the cell, part
is outside — and that makes it difficult for
scientists like Dr. Lupas to determine the
shape of the protein in the lab. Even after a
decade, he could not pinpoint the shape.
With AlphaFold, he cracked the problem
in half an hour.
If these methods continue to improve, hesaid, they could be a particularly useful way
of determining whether a new virus could
be treated with a cocktail of existing drugs.
“We could start screening every com-
pound that is licensed for use in humans,”
Dr. Lupas said. “We could face the next pan-
demic with the drugs we already have.”
During the current pandemic, a simpler
form of artificial intelligence proved helpful
in some cases. A system built by another
London company, BenevolentAI, helped
pinpoint an existing drug, baricitinib, that
could be used to treat seriously ill Covid-19
patients. Researchers have now completed
a clinical trial, though the results have not
yet been released.
As researchers continue to improve the
technology, AlphaFold could further accel-
erate this kind of drug repurposing, as well
as the development of entirely new vac-
cines, especially if we encounter a virus that
is even less understood than Covid-19 is.
David Baker, the director of the Institute
for Protein Design at the University of
Washington, who has been using similarcomputer technology to design anti-corona-
virus drugs, said DeepMind’s methods
could accelerate that work.
“We were able to design coronavirus-
neutralizing proteins in several months,” he
said. “But our goal is to do this kind of thing
in a couple of weeks.”
Still, development speed must contend
with other issues, like large clinical trials,
said Dr. Vincent Marconi, a researcher at
Emory University in Atlanta who helped
lead the baricitinib trial. “That takes time,”
he said.
But DeepMind’s methods could be a way
of determining whether a clinical trial will
fail because of toxic reactions or other prob-
lems, at least in some cases.
Demis Hassabis, DeepMind’s chief exec-
utive and co-founder, said that the company
planned to publish details describing its
work, but that it was unlikely to happen un-
til sometime next year. He also said the
company was exploring ways of sharing the
technology itself with other scientists.
DeepMind is a research lab. It does not
sell products directly to other labs or busi-
nesses. But it could work with other compa-
nies to share access to its technology over
the internet.
The lab’s biggest breakthroughs in the
past have involved games. It built systems
that surpassed human performance on the
ancient strategy game Go and the popular
video game StarCraft — enormously techni-
cal achievements with no practical applica-
tion. Now, the DeepMind team are eager to
push their artificial intelligence technology
into the real world.
“We don’t want to be a leader board com-
pany,” Dr. Jumper said. “We want real bio-
logical relevance.”Artificial Intelligence Cracks a Mystery About Proteins
A lab’s computer system may
help scientists understand
diseases and design drugs.
DEEPMINDScientists have long
sought a better way to
predict the shape of a
twisting string of amino
acids, which defines
what a protein can do.By CADE METZThe DeepMind team includes,
from left, Demis Hassabis, the
lab’s chief executive; John
Jumper, the team’s leader; and
Kathryn Tunyasuvunakool, a
research scientist.