November 30, 2020

DeepMind's AI networks have spent the past few years destroying human players in chess, Go, and classic video games. Now, they're ready to help humans out.

On Monday, DeepMind revealed its AI system AlphaFold had cracked a 50-year-old biological challenge, accurately predicting how proteins within the human body fold into 3D shapes based on their DNA sequences. Those shapes are key in determining how a protein works, and in turn pivotal to figuring out how to treat diseases that involve those proteins, The Guardian explains.

Proteins, which are sequences of amino acids within living creatures, can "bend into a mind-boggling variety of shapes," The Guardian writes. It takes about a year and costs around $120,000 to identify a single protein's shape using the most common method, known as X-ray crystallography, Fortune reports. DeepMind had AlphaFold study 170,000 protein sequences and shapes that had already been identified, and after a few weeks, AlphaFold was ready to face off against other computer-based protein structure predictors in an international competition called CASP.

When asked to extrapolate 100 protein shapes from their amino acid sequences, AlphaFold beat out every other program in CASP and produced results that rivaled lab methods. It predicted a protein's structure within an atom's width of accuracy in two-thirds of those proteins, and was "highly accurate" in the other third, per Fortune. It also only took a few days to identify each protein.

CASP co-founder John Moult called AlphaFold's results a "big deal," telling Nature that "in some sense the problem is solved." If scientists can more quickly figure out a protein's shape, they can find out how it affects other cells — for example, discovering how COVID-19's spike proteins latch onto host cells helped scientists develop vaccines that reduce transmission. DeepMind CEO Demis Hassabis said the company is working on how to share AlphaFold with researchers, and that some scientists have already started using it on vexing protein analyses of their own. Kathryn Krawczyk

April 26, 2019

NASA is concocting a faux asteroid to prepare for a hypothetical apocalyptic threat.

Researchers at NASA and FEMA recently announced plans to partner with NASA's Planetary Defense Coordinator Office, the European Space Agency's Space Situational Awareness-NEO Segment and the International Asteroid Warning Network to test-run the impact of extraterrestrial objects near Earth.

While it's highly unlikely for a UFO to come hurtling toward our planet, scientists do want to prepare for a possible asteroid, reports Gizmodo. Now, the agencies are developing a "realistic— but fictional " exercise that would help experts react in case of such a disaster.

In the simulation, astronomers track the asteroid until they determine there is a 1 percent chance it will collide with the Earth — That's the threshold at which international organizations have agreed they'd need to spring into action. Afterward, scientists measure the "risk corridor," to narrow down the regions of the planet that are most likely to be hit.

"These exercises have really helped us in the planetary defense community to understand what our colleagues on the disaster management side need to know," said Lindley Johnson, NASA's Planetary Defense Officer. "This exercise will help us develop more effective communications with each other and with our governments." Tatyana Bellamy-Walker

January 17, 2019

Scientists have long being trying to determine just how old the rings of Saturn are — did they form at the same time as the planet, 4.5 billion years ago, or are they younger, the result of a moon or comet being pulverized by Saturn's gravitational pull?

NASA's Cassini probe provided the answer. Before it dove into Saturn's atmosphere in 2017, ending its exploration of the planet, Cassini sent back its final pieces of data. The satellite flew between the rings multiple times, and found their mass is 20 times smaller than previous estimates, only about two-fifths the mass of Saturn's moon Mimas. With that information, as well as knowing the proportion of dust in the rings and the rate that dust is added, scientists were able to determine that Saturn's rings could be as young as 10 million years old but no more than 100 million years old.

Looking at the big picture that is the Solar System, this is considered "yesterday," Luciano Iess of Sapienza University in Rome told BBC News. Last month, a group of scientists determined that every 30 minutes, enough ring particles are falling onto Saturn to fill an Olympic-sized pool. Dr. Tom Stallard of Leicester University in the United Kingdom told BBC News the rings will likely disappear in "at most 100 million years," and 50 to 100 million years ago, the rings would have been "even bigger and even brighter" than they are today. Catherine Garcia

November 26, 2018

On Monday, NASA engineers at the Jet Propulsion Lab will be on pins and needles, waiting to see if the InSight spacecraft touches down safely on Mars.

InSight was launched seven months ago, traveling 301,223,981 miles and reaching a top speed of 6,200 mph. This is NASA's first mission to study the deep interior of Mars, and the craft's landing will be tricky. InSight will enter the planet's atmosphere at hypersonic speed, and must slow down quickly in order to make a gentle landing. If all goes according to plan, InSight will touch down at noon PST, with signals reaching engineers back on Earth eight minutes later.

"We've studied Mars from orbit and from the surface since 1965, learning about its weather, atmosphere, geology, and surface chemistry," said Lori Glaze, acting director of the Planetary Science Division in NASA's Science Mission Directorate. "Now we finally will explore inside Mars and deepen our understanding of our terrestrial neighbor as NASA prepares to send human explorers deeper in to the solar system."

The Jet Propulsion Lab's entry, descent, and landing team pre-programmed all stages of InSight's landing, taking into consideration weather reports sent from NASA's Mars Reconnaissance Orbiter, and they might have to slightly tweak the algorithm right before InSight starts its descent. NASA says it took more than a decade to build InSight, and should it land successfully on Mars, it will take two to three months for InSight's robotic arm to set up and calibrate all of the mission's instruments. Catherine Garcia

September 24, 2018

Researchers have found that after only 10 minutes of light exercise, there is enhanced communication between the regions of the brain that store and recall memories.

Scientists from the University of California, Irvine, had 36 healthy volunteers in their early 20s exercise for 10 minutes, doing light activity like yoga or walking. The volunteers then took a memory test, which was repeated later without exercise. The researchers asked 16 of the volunteers to take the test again, with some exercising first and others resting. While studying their brain activity, it was discovered that those who exercised had increased activity between the hippocampus and cortical brain regions, all of which are associated with memory.

The results were published in the journal Proceedings of the National Academy of Sciences, and the researchers concluded that volunteers who exercised had an easier time distinguishing between different memories. Michael Yassa, a neuroscientist at UCI and project co-leader, told The Guardian that the amount of necessary exercise is dependent on a person's age, mobility level, and other lifestyle factors, and for many, taking a leisurely stroll is enough. Catherine Garcia

September 15, 2018

California's state government will work with a company called Planet Labs to build and launch a climate change research satellite capable of tracking the sources of pollution, California Gov. Jerry Brown (D) said Friday.

"We're going to launch our own satellite — our own damn satellite to figure out where the pollution is and how we're going to end it," Brown said at the Global Climate Action Summit in San Francisco. "This groundbreaking initiative will help governments, businesses, and landowners pinpoint — and stop — destructive emissions with unprecedented precision, on a scale that's never been done before."

Brown did not say when the satellite would be ready or how much it might cost. A statement from his office said the project "has the potential to deliver global emission reductions equivalent to 1,000 million metric tons of carbon dioxide annually — or removing 200 million vehicles from roads every year."

The day before, at the same event, Brown argued environmental issues would be a major black mark on President Trump's legacy. "When Trump says, in effect, 'We like more methane going into the air,' that is highly destructive, very highly destructive," he said. "So I think he'll be remembered [as] he is now — I don't know: liar, criminal, fool, pick your choice." Bonnie Kristian

August 30, 2018

Everyday medicine is getting a technical transformation.

The first "digital pill" with an implanted sensor is ready to make its debut, Stat reports. These high-tech capabilities only come in the antipsychotic drug Abilify MyCite right now, but the development could soon provide unprecedented insight into how medicine works.

Abilify MyCite, a treatment for bipolar disorder and schizophrenia, was first granted FDA approval last year, per its maker Otsuka. Its tiny internal sensor knows when it touches stomach fluid and sends details to the patient's wearable MyCite patch. Patients can then look at ingestion data on an app and input how they're feeling, and doctors can view the information online.

The generic of Abilify costs around $700 per month, and MyCite will run about $1,650 for takers without insurance, Stat says, so only a handful of Medicaid users will get to try the high-tech pill for now. Otsuka say it's capitalizing on this small-scale debut to learn from how the doctors and patients use the sensor. Results could inform the development of more precise mental illness medications, the company says, and perhaps transform medication as we know it. Read more at Stat. Kathryn Krawczyk

February 20, 2018

Our eyeballs apparently contain information that could revolutionize cardiovascular medicine.

Artificial intelligence software developed by Google in conjunction with its biotech subsidiary company Verily can scan retinal images to predict heart disease at nearly the same accuracy rate as a traditional blood test, United Press International reports. The findings, published Monday in the journal Nature Biomedical Engineering, explain that Google's AI makes its predictions by examining images of the back of a patient's eye in order to develop a profile of the patient, including several characteristics that could determine cardiovascular risk.

From the retinal images, Google's AI can determine within impressive degrees of accuracy a patient's age, gender, blood pressure, and smoking status, as well as even the past occurrence of major cardiovascular events, The Verge explains. The program taught itself how to analyze eyeballs after using machine learning techniques to pore over more than 284,000 retinal images; while studying, the AI used what UPI describes as a visual "heatmap" to learn which parts of the eye's anatomy contained certain predictive factors. The AI eventually learned, for example, that to analyze a patient's blood pressure, it was prudent to examine the blood vessels in the eye.

To test its capabilities, researchers sicced the AI on two patient pools, totaling more than 13,000 patients. The AI made correct predictions on the future risk of heart disease in 70 percent of cases — nearly the same accuracy rate as the blood-test method doctors traditionally use, which has a 72 percent accuracy rate.

Harlan M. Krumholz, the director of Yale's Center for Outcomes Research and Evaluation, predicted that the findings of Google's AI show that machine learning and artificial intelligence will "more precisely hone our understanding of disease and individuals," helping physicians "understand these processes and diagnoses in ways that we haven't been able to before." Read the full study here. Kelly O'Meara Morales

See More Speed Reads