Making New Drugs With a Dose of Artificial Intelligence
Every two years, hundreds of scientists enter a global competition. Tackling a biological puzzle they call "the protein folding problem," they try to predict the three-dimensional shape of proteins in the human body. No one knows how to solve the problem. Even the winners only chip away at it. But a solution could streamline the way scientists create new medicines and fight disease.
Mohammed AlQuraishi, a biologist who has dedicated his career to this kind of research, flew in early December to Cancun, Mexico, where academics were gathering to discuss the results of the latest contest. As he checked into his hotel, a five-star resort on the Caribbean, he was consumed by melancholy. The contest, the Critical Assessment of Structure Prediction, was not won by academics. It was won by DeepMind, the artificial intelligence lab owned by Google's parent company. "I was surprised and deflated," said Dr. AlQuraishi, a researcher at Harvard Medical School. "They were way out in front of everyone else."
[...] "It is not that machines are going to replace chemists," said Derek Lowe, a longtime drug discovery researcher and the author of In the Pipeline, a widely read blog dedicated to drug discovery. "It's that the chemists who use machines will replace those that don't."
After the conference in Cancun, Dr. AlQuraishi described his experience in a blog post. The melancholy he felt after losing to DeepMind gave way to what he called "a more rational assessment of the value of scientific progress." But he strongly criticized big pharmaceutical companies like Merck and Novartis, as well as his academic community, for not keeping pace.
[For those who might not be aware, the Derek Lowe mentioned above is the author of the Things I Won't Work With blog. If you want to read about things that burn, go BOOM, or otherwise wreak havoc you'd be hard pressed to find a more entertaining source. --martyb]
(Score: 2) by opinionated_science on Wednesday February 13 2019, @12:03PM
All these algorithms are brute force - with constraints from existing protein structures (xray, nmr, etc).
"Deepmind" is a marketing division of Google to buy their stuff - impressive they pointed it at the problem - not sure how much it helps.
"deepmind" is *really* useful if you have a mountain of classified (i.e. pre tagged) data, but worthless in telling you *how* to classify.
(Score: 2) by Spamalope on Wednesday February 13 2019, @02:25PM (1 child)
If you want to read about things that burn, go BOOM
Definitely read the FOOF post, along with some of the comments.
(Score: 0) by Anonymous Coward on Wednesday February 13 2019, @03:01PM
It's amazing how easy it is to kill a living creature with some molecules.
(Score: 2) by hendrikboom on Thursday February 14 2019, @09:37PM
It is already known that folding doesn't depend on the base sequence alone. There are other chemicals, called "chaperones" that affect how a protein will fold.
Pure brute-force chemical-bond computation will have to calculate the effects of the (unknown) chaperones on the folding process.
So what was the AI trained on? Known proteins and the way they folded? Those training data will already contain the consequences of the (unknown) chaperones. This is an advantage. It enables the AI to guess straight through to the result without yielding any insight into the process.