From Atoms To Words #7: From Serendipity to Data-Driven Discoveries

Welcome to?From Atoms to Words ! This month, we explore the shift in science, from serendipity to data-driven discoveries. Is there still room for curiosity and intuition? Certainly. Think of graphene and Scotch Tape. And yet, R&D is moving to a more data-centric approach, deploying computational techniques, from (quantum) simulations to machine learning. The quintessential example of this? AlphaFold's landmark achievement in protein structure prediction. Ready to dive in? Let's go!

Curiosity, Ingenuity, Persistence – Andre Geim’s Random Walk to the Discovery of Graphene

Everyone is familiar with the pencil, that marvelous instrument wielded by poets to scribble, and sometimes erase, their most regrettable verses.

? Pencils are made of graphite—a form of carbon arranged in parallel sheets, weakly bound to each other via van der Waals interactions.

Subject graphite to the right conditions, and it transforms into a diamond. Interestingly, given enough time—about 1 billion years—a diamond will revert to graphite.

?? So much for diamonds being forever, huh?

Now, if you could manipulate graphite in just the right way, you'd isolate the thinnest and lightest material known to humankind—a material that is roughly 100 times stronger than steel and an unparalleled conductor of both electricity and heat.

? All this wonder is packed into a single layer of carbon atoms. That wonder, dear reader, is graphene.

You might be thinking, "It's just pencil, dude!" How hard can it be to get graphene out of graphite?

It's not that hard. And yet, it literally took centuries. That is, until Andre Geim and Kostya Novoselov's seismic paper in 2004.

?? So, grab your notebook and your pencil (ok, bad joke) and let's journey through a brief history of graphene:

?? What do frogs and geckos have to do with the discovery of graphene?

?? Why running experiments on Friday nights may lead you to the Nobel prize?

?? How did a piece of Scotch Tape revolutionize graphene research?

Curious?

?? Head to Curiosity, Ingenuity, Persistence – Andre Geim’s Random Walk to the Discovery of Graphene

Bridging Theory and Experiment: 14 Reasons Chemical Simulations Stand as the Third Pillar of R&D

Ah, chemical simulations. Every journey has a beginning, and mine was during my undergrad years, exploring the world of bioinorganic and physical chemistry with nothing but precision scales and spectrophotometers. ??

But then my experimental research threw me a curveball: quantum chemistry. ?

Suddenly, I found myself working with chemical simulations on a computer cluster. What began as an auxiliary technique to support my lab experiments quickly turned into an essential part of my research toolkit.

?? And I'm not alone; a growing number of researchers in both academia and industry are finding simulations indispensable.

Gone are the days when computational chemistry was an arcane art mastered only by a select few.

?? At my gig at Quantistry , I'm seeing it happen in real-time—from automakers to aerospace giants, industrial players are integrating these simulations into their R&D like never before.

And if you're impressed now, just wait until quantum computing and machine learning really get in the game.

??So, do I think simulations are turning into the third pillar of science?

You bet. And it's not just a gut instinct. Simulations are gaining momentum. But why this shift?

?? Why are today's scientists leaning more into the computational domain?

?? What's the appeal of chemical simulations for R&D?

?? How can simulations complement and support experiments?

?? Find my personal answers at Bridging Theory and Experiment: 14 Reasons Chemical Simulations Stand as the Third Pillar of R&D

60 Years in the Making: AlphaFold's Historical Breakthrough in Protein Structure Prediction

?? Why is protein structure prediction considered a mission impossible? And why is AlphaFold such a game changer?

Cast your mind back to the 1960s, a time of great scientific fervor following the discovery of the DNA double helix. ??

There was hopeful anticipation that protein structures would also exhibit some elegant internal pattern and regularity. However, the unveiling of myoglobin's structure threw a wrench in those assumptions, presenting helices haphazardly packed together.

?? This marked the starting point of the protein folding problem.

So, why is predicting how a protein folds from its amino-acid sequence such an enormous challenge?

The thing is, an unfolded protein possesses an astronomical number of potential configurations. Literally. ?

Consider a protein chain with 100 residues. This has 99 peptide bonds, which result in 198 distinct bond angles. Now, imagine that each of these angles can take on one of three possible conformations. This amounts to 3^198 potential possibilities!

? If the protein were to explore all of them, one by one, the folding would take longer than the age of the universe.Astonishingly, most proteins manage to fold into their correct structures within milliseconds or microseconds.

Faced with such hellish combinatorics, early attempts at predicting protein structure relied on databases and computational methods, such as force fields and Monte Carlo.

? In 1994, a maverick by the name of John Moult stepped onto the scene with his Critical Assessment of Techniques for Protein Structure Prediction, or CASP for short. This biennial competition still invites research groups to take part in a blind test—predicting protein structures solely from the amino-acid sequence, without any knowledge of the experimental results.

For grueling decades, progress at CASP eked forward, each advance coming at a painstaking crawl.

That was, until 2020.

?? A staggering 60 years after the first experimental protein structures, AlphaFold burst onto the scene in a blaze of glory. Harnessing the power of machine learning, it delivered results so astonishing, they left the scientific community agog. ??

?? How effectively does AlphaFold predict protein structures?

?? What's under the hood of AlphaFold?

?? What implications does it have for scientific and medical research?

?? Dive deeper, explore references, and perhaps chuckle at the usual bad puns at 60 Years in the Making: AlphaFold's Historical Breakthrough in Protein Structure Prediction

+3 Bonus Stories

?? Large Language Models for Chemistry: Is the Beginning of a New Era? [Read more ]

?? The Evolution of Quantum Chemistry: From Pencil and Paper to Quantum Computing. [Read more ]

?? Computational Chemistry 2043: A Quantum Peep into the Future. [Read more ]

Did you find this newsletter helpful or insightful?

Subscribe ?to?#FromAtomsToWords ?to receive future stories about quantum chemistry, simulations, machine learning and the world around it. Let me know your comments or suggestions below, and thank you for reading!

???Read previous issues of From Atoms To Words

?#ResearchAndDevelopment #ComputationalChemistry #ProteinStructurePrediction #GrapheneDiscovery #ChemicalSimulations #FromAtomsToWords #AlphaFold