[ad_1]
Unbeknownst to them, at DeepMind, a large scientific paper about their system was already under review. Nature, according to John Jumper, director of AlphaFold. DeepMind had submitted their entries to Nature on May 11.
At the time, scientists didn’t know much about DeepMind’s time. That changed three days after Baker’s first president’s inauguration, on June 18, when DeepMind CEO Demis Hassabis took to Twitter. “We’ve been working hard on our full (reviewed) papers and what we have and giving us access to AlphaFold for scientists,” he wrote. “Soon!”
On July 15, the same day that Baker’s RoseTTAFold paper was published, Nature released DeepMind unaltered but peer-reviewed AlphaFold2 manuscripts. Immediately, DeepMind created the AlphaFold2 code free access on github. And a week later, the group was released An very large database of the 350,000 proteins predicted by their method. The predictive tool for protein, as well as its predicted quantity, was finally in the hands of scientists.
According to Jumper, there is a reason to fight over why the paper and DeepMind code were not released until more than seven months later the CASP show states: “We were not ready to open the source or publish this in detail that day,” he said. . When the newspaper was released in May, and the group was working on a review, Jumper said it had made efforts to publish the paper as soon as possible. “We’ve been as sincere as possible,” he says.
DeepMind team manuscripts were published via NatureA Quick Review Article, which the magazine uses extensively on Covid-19 papers. In his remarks to WIRED, the prophet of Nature he wrote that the approach “is a work in progress for our writers and readers, in order to make their peers interested and time-conscious re-evaluating the availability of fast-track research.”
Jumper and Pushmeet Kohli, leader of the scientific team DeepMind, are not sure if Baker’s paper was fulfilled in their time. Nature printing. “From our point of view, we supported and submitted the paper in May, so it was in our hands, in a way,” Kohli said.
But CASP co-founder Moult believes the University of Washington team work may have helped DeepMind scientists convince their parent company that their research will be available on a regular basis. “My experience with knowing them – they are well-known scientists – is that they want to be as open as possible,” says Moult. “There are some problems there, because it’s a business, and in the end it has to make money in some way.” The company with DeepMind, Alphabet, has the fourth largest market in the world.
Hassabis is known for the release of AlphaFold2 as a benefit to both scientists and Alumni. “All of this is open science and we are delivering this to the public, no cables, system, rules, and database,” he said in an interview with WIRED. Asked if there was a discussion on confidentiality for commercial reasons, he said, “It’s a good question how we offer profits. Price can be offered in a variety of ways, isn’t it? One is obviously commercial, but there is also popularity.”
Baker is quick to thank the DeepMind team for fully reviewing their release of papers and codes. In other words, he says, RoseTTAFold was a wall against the possibility that DeepMind would not act in accordance with science. “If only they had not been enlightened and decided to do so [release the code], then the world would be a better place for him to build, ”he says.
[ad_2]
Source link
