Once a week.
Subscribe to our weekly newsletter.
DeepMind AI solves 50-year-old biology problem in breakthrough advance
The Google-owned company developed a system that can reliably predict the 3D shapes of proteins.
- Scientists have long been puzzled by how specific chains of amino acids go on to form three-dimensional proteins.
- DeepMind developed a system that's able to predict "protein folding" in a fraction of the time of human experiments, and with unprecedented accuracy.
- The achievement could greatly improve drug research and development, as well as bioengineering pursuits.
In 1994, a group of scientists created a competition to solve one of the most perplexing problems in biology: how do proteins fold themselves into 3D shapes, which then carry out fundamental processes within living organisms?
The answer to this 50-year-old question could revolutionize many scientific pursuits, from accelerating and improving drug development, to creating better biofuels. But the competition, called Critical Assessment of Protein Structure Prediction (CASP), went decades without a solution.
Then artificial intelligence got into the mix.
DeepMind, a U.K.-based AI company, essentially solved the long-standing problem in the most recent competition, CASP14. The company outperformed the other teams by magnitudes, predicting the shapes of proteins with accuracy rates never before achieved by humans.
"This is a big deal," John Moult, a computational biologist who co-founded CASP, told Nature. "In some sense the problem is solved."
In the biennial competition, teams analyze around 100 proteins with the goal of predicting their eventual 3D shape. A protein's shape determines its function. For example, a protein can become an antibody that binds to foreign particles to protect, an enzyme that carries out chemical reactions, or a structural component that supports cells.
Proteins start as a string of hundreds of amino acids. Within a protein, pairs of amino acids can interact in numerous ways, and these particular interactions determine the final shape of the protein. But given the sheer number of possible interactions, it's incredibly difficult to predict a protein's physical shape. Difficult, but not impossible.
Since CASP began, scientists have been able to predict the shape of some simple proteins with reasonable accuracy. CASP is able to verify the accuracy of these predictions by comparing them to the actual shape of proteins, which it obtains through the unpublished results of lab experiments.
But these experiments are difficult, often taking months or years of hard work. The shapes of some proteins have eluded scientists for decades. As such, it's hard to overstate the value of having an AI that's able to churn out this work in just hours, or even minutes.
In 2018, DeepMind, which was acquired by Google in 2014, startled the scientific community when its AlphaFold algorithm won the CASP13 contest. AlphaFold was able to predict protein shapes by "training" itself on vast amounts of data on known amino acid strings and their corresponding protein shapes.
In other words, AlphaFold learned that particular amino acid configurations—say, distances between pairs, angles between chemical bonds—signaled that the protein would likely take a particular shape. AlphaFold then used these insights to predict the shapes of unmapped proteins. AlphaFold's performance in the 2018 contest was impressive, but not reliable enough to consider the problem of "protein folding" solved.
In the latest contest, DeepMind used an updated version of AlphaFold. It combines the previous deep-learning strategy with a new "attention algorithm" that accounts for physical and geometric factors. Here's how DeepMind describes it:
"A folded protein can be thought of as a 'spatial graph,' where residues are the nodes and edges connect the residues in close proximity. This graph is important for understanding the physical interactions within proteins, as well as their evolutionary history."
"For the latest version of AlphaFold, used at CASP14, we created an attention-based neural network system, trained end-to-end, that attempts to interpret the structure of this graph, while reasoning over the implicit graph that it's building. It uses evolutionarily related sequences, multiple sequence alignment (MSA), and a representation of amino acid residue pairs to refine this graph."
CASP measures prediction accuracy through the "Global Distance Test (GDT)", which ranges from 0-100. The new version of AlphaFold scored a median of 92.4 GDT for all targets.
Given that the specific ways in which proteins take shape can shed light on how diseases form, AlphaFold could greatly accelerate disease research and drug development. And while it's too late for the system to help with COVID-19, DeepMind says that protein structure prediction could be "useful in future pandemic response efforts."
Still, scientists have much to learn about predicting protein structures, and while AlphaFold has proven faster and more accurate than human experiments, the system isn't 100 percent accurate. But DeepMind's achievement signals that AI may become a surprisingly powerful tool in unlocking key mysteries in biology and beyond.
"For all of us working on computational and machine learning methods in science, systems like AlphaFold demonstrate the stunning potential for AI as a tool to aid fundamental discovery," DeepMind wrote. "Just as 50 years ago Anfinsen laid out a challenge far beyond science's reach at the time, there are many aspects of our universe that remain unknown. The progress announced today gives us further confidence that AI will become one of humanity's most useful tools in expanding the frontiers of scientific knowledge, and we're looking forward to the many years of hard work and discovery ahead!"
- A new day is coming in healthcare, where AI will help diagnose and ... ›
- Listen to New Google AI Program Talk Like a Human and Write Music ›
- Google's AI Learns Betrayal and "Aggressive" Actions Pay Off - Big ... ›
Evolution doesn't clean up after itself very well.
- An evolutionary biologist got people swapping ideas about our lingering vestigia.
- Basically, this is the stuff that served some evolutionary purpose at some point, but now is kind of, well, extra.
- Here are the six traits that inaugurated the fun.
The plica semilunaris<img class="rm-lazyloadable-image rm-shortcode" type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8xOTA5NjgwMS9vcmlnaW4ucG5nIiwiZXhwaXJlc19hdCI6MTY3NDg5NTg1NX0.kdBYMvaEzvCiJjcLEPgnjII_KVtT9RMEwJFuXB68D8Q/img.png?width=980" id="59914" width="429" height="350" data-rm-shortcode-id="b11e4be64c5e1f58bf4417d8548bedc7" data-rm-shortcode-name="rebelmouse-image" />
The human eye in alarming detail. Image source: Henry Gray / Wikimedia commons<p>At the inner corner of our eyes, closest to the nasal ridge, is that little pink thing, which is probably what most of us call it, called the caruncula. Next to it is the plica semilunairs, and it's what's left of a third eyelid that used to — ready for this? — blink horizontally. It's supposed to have offered protection for our eyes, and some birds, reptiles, and fish have such a thing.</p>
Palmaris longus<img class="rm-lazyloadable-image rm-shortcode" type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8xOTA5NjgwNy9vcmlnaW4uanBnIiwiZXhwaXJlc19hdCI6MTYzMzQ1NjUwMn0.dVor41tO_NeLkGY9Tx46SwqhSVaA8HZQmQAp532xLxA/img.jpg?width=980" id="879be" width="1920" height="2560" data-rm-shortcode-id="4089a32ea9fbb1a0281db14332583ccd" data-rm-shortcode-name="rebelmouse-image" />
Palmaris longus muscle. Image source: Wikimedia commons<p> We don't have much need these days, at least most of us, to navigate from tree branch to tree branch. Still, about 86 percent of us still have the wrist muscle that used to help us do it. To see if you have it, place the back of you hand on a flat surface and touch your thumb to your pinkie. If you have a muscle that becomes visible in your wrist, that's the palmaris longus. If you don't, consider yourself more evolved (just joking).</p>
Darwin's tubercle<img class="rm-lazyloadable-image rm-shortcode" type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8xOTA5NjgxMi9vcmlnaW4uanBnIiwiZXhwaXJlc19hdCI6MTY0ODUyNjA1MX0.8RuU-OSRf92wQpaPPJtvFreOVvicEwn39_jnbegiUOk/img.jpg?width=980" id="687a0" width="819" height="1072" data-rm-shortcode-id="ff5edf0a698e0681d11efde1d7872958" data-rm-shortcode-name="rebelmouse-image" />
Darwin's tubercle. Image source: Wikimedia commons<p> Yes, maybe the shell of you ear does feel like a dried apricot. Maybe not. But there's a ridge in that swirly structure that's a muscle which allowed us, at one point, to move our ears in the direction of interesting sounds. These days, we just turn our heads, but there it is.</p>
Goosebumps<img class="rm-lazyloadable-image rm-shortcode" type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8xOTA5NzMxNC9vcmlnaW4uanBnIiwiZXhwaXJlc19hdCI6MTYyNzEyNTc2Nn0.aVMa5fsKgiabW5vkr7BOvm2pmNKbLJF_50bwvd4aRo4/img.jpg?width=980" id="d8420" width="1440" height="960" data-rm-shortcode-id="8827e55511c8c3aed8c36d21b6541dbd" data-rm-shortcode-name="rebelmouse-image" />
Goosebumps. Photo credit: Tyler Olson via Shutterstock<p>It's not entirely clear what purpose made goosebumps worth retaining evolutionarily, but there are two circumstances in which they appear: fear and cold. For fear, they may have been a way of making body hair stand up so we'd appear larger to predators, much the way a cat's tail puffs up — numerous creatures exaggerate their size when threatened. In the cold, they may have trapped additional heat for warmth.</p>
Tailbone<img class="rm-lazyloadable-image rm-shortcode" type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8xOTA5NzMxNi9vcmlnaW4uanBnIiwiZXhwaXJlc19hdCI6MTY3MzQwMjc3N30.nBGAfc_O9sgyK_lOUo_MHzP1vK-9kJpohLlj9ax1P8s/img.jpg?width=980" id="9a2f6" width="1440" height="1440" data-rm-shortcode-id="4fe28368d2ed6a91a4c928d4254cc02a" data-rm-shortcode-name="rebelmouse-image" />
Image source: Decade3d-anatomy online via Shutterstock<p>Way back, we had tails that probably helped us balance upright, and was useful moving through trees. We still have the stump of one when we're embryos, from 4–6 weeks, and then the body mostly dissolves it during Weeks 6–8. What's left is the coccyx.</p>
The palmar grasp reflex<img class="rm-lazyloadable-image rm-shortcode" type="lazy-image" data-runner-src="https://assets.rebelmouse.io/eyJhbGciOiJIUzI1NiIsInR5cCI6IkpXVCJ9.eyJpbWFnZSI6Imh0dHBzOi8vYXNzZXRzLnJibC5tcy8xOTA5NzMyMC9vcmlnaW4uanBnIiwiZXhwaXJlc19hdCI6MTYzNjY0MDY5NX0.OSwReKLmNZkbAS12-AvRaxgCM7zyukjQUaG4vmhxTtM/img.jpg?width=980" id="8804c" width="1440" height="960" data-rm-shortcode-id="67542ee1c5a85807b0a7e63399e44575" data-rm-shortcode-name="rebelmouse-image" />
Palmar reflex activated! Photo credit: Raul Luna on Flickr<p> You've probably seen how non-human primate babies grab onto their parents' hands to be carried around. We used to do this, too. So still, if you touch your finger to a baby's palm, or if you touch the sole of their foot, the palmar grasp reflex will cause the hand or foot to try and close around your finger.</p>
Other people's suggestions<p>Amir's followers dove right in, offering both cool and questionable additions to her list. </p>
Fangs?<blockquote class="twitter-tweet" data-conversation="none" data-lang="en"><p lang="en" dir="ltr">Lower mouth plate behind your teeth. Some have protruding bone under the skin which is a throw back to large fangs. Almost like an upsidedown Sabre Tooth.</p>— neil crud (@neilcrud66) <a href="https://twitter.com/neilcrud66/status/1085606005000601600?ref_src=twsrc%5Etfw">January 16, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Hiccups<blockquote class="twitter-tweet" data-conversation="none" data-lang="en"><p lang="en" dir="ltr">Sure: <a href="https://t.co/DjMZB1XidG">https://t.co/DjMZB1XidG</a></p>— Stephen Roughley (@SteBobRoughley) <a href="https://twitter.com/SteBobRoughley/status/1085529239556968448?ref_src=twsrc%5Etfw">January 16, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Hypnic jerk as you fall asleep<blockquote class="twitter-tweet" data-conversation="none" data-lang="en"><p lang="en" dir="ltr">What about when you “jump” just as you’re drifting off to sleep, I heard that was a reflex to prevent falling from heights.</p>— Bann face (@thebanns) <a href="https://twitter.com/thebanns/status/1085554171879788545?ref_src=twsrc%5Etfw">January 16, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script> <p> This thing, often called the "alpha jerk" as you drop into alpha sleep, is properly called the hypnic jerk,. It may actually be a carryover from our arboreal days. The <a href="https://www.livescience.com/39225-why-people-twitch-falling-asleep.html" target="_blank" data-vivaldi-spatnav-clickable="1">hypothesis</a> is that you suddenly jerk awake to avoid falling out of your tree.</p>
Nails screeching on a blackboard response?<blockquote class="twitter-tweet" data-conversation="none" data-lang="en"><p lang="en" dir="ltr">Everyone hate the sound of fingernails on a blackboard. It's _speculated_ that this is a vestigial wiring in our head, because the sound is similar to the shrill warning call of a chimp. <a href="https://t.co/ReyZBy6XNN">https://t.co/ReyZBy6XNN</a></p>— Pet Rock (@eclogiter) <a href="https://twitter.com/eclogiter/status/1085587006258888706?ref_src=twsrc%5Etfw">January 16, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Ear hair<blockquote class="twitter-tweet" data-conversation="none" data-lang="en"><p lang="en" dir="ltr">Ok what is Hair in the ears for? I think cuz as we get older it filters out the BS.</p>— Sarah21 (@mimix3) <a href="https://twitter.com/mimix3/status/1085684393593561088?ref_src=twsrc%5Etfw">January 16, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Nervous laughter<blockquote class="twitter-tweet" data-lang="en"><p lang="en" dir="ltr">You may be onto something. Tooth-bearing with the jaw clenched is generally recognized as a signal of submission or non-threatening in primates. Involuntary smiling or laughing in tense situations might have signaled that you weren’t a threat.</p>— Jager Tusk (@JagerTusk) <a href="https://twitter.com/JagerTusk/status/1085316201104912384?ref_src=twsrc%5Etfw">January 15, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Um, yipes.<blockquote class="twitter-tweet" data-conversation="none" data-lang="en"><p lang="en" dir="ltr">Sometimes it feels like my big toe should be on the side of my foot, was that ever a thing?</p>— B033? K@($ (@whimbrel17) <a href="https://twitter.com/whimbrel17/status/1085559016011563009?ref_src=twsrc%5Etfw">January 16, 2019</a></blockquote> <script async src="https://platform.twitter.com/widgets.js" charset="utf-8"></script>
Are we enslaved by the finer things in life?
- The Roman writer, Tacitus, argued that the Roman Empire was built by enslaving conquered people who became accustomed to fine living and luxury.
- Technology today has become so essential to our daily lives that it seems impossible to break free of it. It's as much a cage as a luxury.
- Being dependent on a thing gives it power over you. To need something or someone is, for better or worse, to limit yourself.
- There was a massive die-off of marine life 359 million years ago, and nobody knows why.
- A new study proposes that the Late Devonian extinction may have been caused by one or more nearby supernovae.
- The supernova hypothesis could be confirmed if scientists can find "the green bananas of the isotope world" in the geologic record.
Fifty years of research on children's toy preferences shows that kids generally prefer toys oriented toward their own gender.