« December 3, 2020 | Main | December 5, 2020 »

December 4, 2020

DeepMind AlphaFold AI wins 'Olympics of protein folding'

From the Verge:

Within any living organism, there are thousands of different proteins, each with its own unique shape.

For decades, the exact formation of those shapes has been a pain for scientists to figure out.

How exactly does a protein, which starts as a string of amino acids, fold itself into the 3D shapes you might recognize from diagrams?

AlphaFold, an AI from DeepMind, may have an answer.

It can predict, with heretofore unseen accuracy, the shape a protein will take.

AlphaFold was put to the test in a global competition called Critical Assessment of protein Structure Prediction, or CASP, which DeepMind CEO Demis Hassabis calls the "Olympics of protein folding" in a video (above).

During the competition, systems like AlphaFold are given the amino acid strings for proteins with shapes that have already been identified through previous experiments but haven't been published yet.

Judges compare the protein shapes produced by the systems with what they know the shapes should be.

At the end of the competition, AlphaFold had the most accurate predictions of any CASP participant in its 25-year history by a wide margin.

Even the predictions that weren't accurate enough to be considered "competitive" with experimental results were only a few atom-widths off.

The complete data still needs to be peer-reviewed and published, but the DeepMind team is excited about the results so far, saying in a blog post that they are "optimistic about the impact AlphaFold can have on biological research and the wider world."

It can take years in the lab for scientists to identify the shapes of individual proteins.

Neural networks like AlphaFold could help speed up biological research and drug development in the future.

The AI method isn't perfect yet, and it won't be taking over for flesh-and-blood researchers anytime soon, but it could be a major step in ongoing scientific advancement.

"What is the protein folding problem?" here.

More from DeepMind here.

December 4, 2020 at 04:01 PM | Permalink | Comments (0)

BE CAREFUL OF LANDSLIDE

1_sEmsoQajkEu9Y1EsHbBeKw-1

Above, a sign in the workshop at the Shenzhen-based hard-tech venture capital firm Haxl8r, now known as HAX.

December 4, 2020 at 02:01 PM | Permalink | Comments (0)

Only God can make a tree

Screen Shot 2020-12-02 at 7.25.36 AM

 

Map shows the area of virgin old-growth forest in the contiguous United States in 1620, 1850, and 1920.

Wait a sec — what's that poem I'm hearing in my head?

[via reddit]

December 4, 2020 at 12:01 PM | Permalink | Comments (1)

World's first digital photograph

Screen Shot 2020-12-02 at 6.59.57 AM

From PetaPixel

The first digital camera, invented in 1975, didn't actually produce the first digital photograph.

The first digital photo (above) actually came almost two decades earlier in 1957 when Russell Kirsch made a 176×176-pixel digital image by using an image scanner invented by his team at NIST to scan a photograph of his three-month-old son Walden Kirsch.

Since each pixel could only show 1 bit of information (black or white), two superimposed binary scans at different thresholds were used to create a composite producing approximate gray levels.

The low resolution was due to the fact that the computer they used wasn't capable of storing more information.

The now iconic image was named one of LIFE magazine's "100 Photographs That Changed the World" in 2003.

The original image is in the Portland (Oregon) Art Museum.

December 4, 2020 at 10:01 AM | Permalink | Comments (0)

Limited-Edition (1) Louvre x Vacheron Constantin

63c0bd3e0ca1bbe4733c93ed1c5ea4a19b4487a8

From Barron's:

Lot 6 in this month's Christie's online auction, "Bid for the Louvre," is dubbed "A Masterpiece on Your Wrist," presenting the opportunity to select a work of art from The Louvre's world-famous collections and have it reproduced on the dial of a one-of-a-kind watch by the master enamellers in Vacheron Constantin's Geneva workshops. 
Created by the brand's Les Cabinotiers custom watchmaking department, the unique piece is estimated to sell for between €100,000 and €300,000 (US$112,660 and US$337,980) with a starting bid of €80,000 (US$90,128).
 
Like all 24 lots in the sale, there is no buyer's premium, with 100% of the proceeds benefiting the Louvre's social and educational programs and the Studio, a new space dedicated to artistic and cultural education that will open next fall.
 
Bidding is open until December 15.
 
Lot 6's winning bidder will walk away with far more than a watch, as the package incorporates an experiential aspect sure to cultivate a heightened appreciation for the Parisian museum's considerable holdings as well as the intricate art of watchmaking. 
 
The experience kicks off with a private visit to the Louvre guided by one of the museum's leading experts, who will enlighten the buyer about its vast holdings of paintings and sculptures.
 
From the Venus de Milo to Delacroix's Liberty Leading the People to Vermeer's The Lacemaker, and countless others, the subject of the watch's dial is solely up to the client’s discretion with the exception of da Vinci's Mona Lisa, the only masterpiece that's off limits.
 
From Paris, the journey continues to Geneva for a tour of Vacheron Constantin's HQ, providing a behind-the-scenes look at the craftsmanship that goes into producing an horological work of art.
 
The client will meet with master watchmakers and designers to discuss their personalized watch and choose from several customization options, including the case metal, personalized engraving, strap, and more.  
But for this particular commission, the real magic happens in the métiers d'art workshops, where a team of master enamellers, engravers, and gem setters ply their age-old decorative crafts. 
 
Depending on the motif selected, the master enameller may recommend miniature enamel painting, a meticulous method known as "Geneva technique" (top), which dates to the 18th century.
 
Using brushes as fine as a single hair, they recreate the original artwork on the dial, which is fired in a kiln with each layer of colored enamel.
 
This painstaking process demands more than just painting skills since the pigments, which can be altered during the firing process, must exactly match the original work of art.
 
Another option is the monochromatic Grisaille (French for "grayness") enamel, a craft developed in the 16th century in France by the Limoges school of enamellers.
 
Aa94e4716c559b2a587209beaab7896986bf504c
 
The technique involves superimposing touches of a rare white enamel — called Limoges white — on a layer of dark enamel coating the gold dial base.
 
The artisan fires the piece in a kiln multiple times, strictly adhering to specified firing times to the nearest second.
 
The watch's transparent case back will reveal the self-winding Calibre 2460 SC fitted with a 22-karat gold oscillating weight sculpted in the shape of the brand's emblematic Maltese cross.
 
Eab5aaebc227e05f7e542c6bcaf1817cba038bf1
 
In keeping with the strict standards of the Hallmark of Geneva, or Geneva Seal, the movement is beautifully hand finished with beveling, circular-graining, and polishing.
 
Once the piece is ready for delivery, the client returns to the Louvre for another after-hours visit and the presentation of the watch next to the subject work of art, accompanied by certificates of authenticity from both Vacheron Constantin and the Louvre Museum.
The online auction is a joint venture with Musée du Louvre and Christie's, with the support of the Hôtel Drouot.
 
As a recent Louvre partner, Vacheron Constantin's contribution is the only horological lot in the mix.
 
"Putting up for auction a Les Cabinotiers timepiece based on a masterpiece, a one-of-a-kind model personalized in accordance with the acquirer's wishes, symbolizes the identity of our maison and its mission to promote the sharing of culture and emotions," said Louis Ferla, Vacheron Constantin CEO, in a news release.

December 4, 2020 at 08:01 AM | Permalink | Comments (0)

« December 3, 2020 | Main | December 5, 2020 »