The new "folding" of protein! DeepMind launches new family member AlphaFold

The first big family in the field of artificial intelligence, the Alpha family from DeepMind, is the favorite thing to do.

All of them are really lucky, and they are expected to be wonderful.

For example, AlphaGo, which was famous in the previous World War, AlphaZero, which followed the arrogant, and AlphaFold, who just joined the family team in the past two days, the Chinese title is Alpha Folding.

In the field of biomedicine, calmness and enthusiasm have always existed together, and the character of the new darling AlphaFold is still unclear, but at least for now, it is bound to trigger a new round of research acceleration in the biomedical field.

"Little Wonderful" Alpha Folding

According to DeepMind, the significance of the Alpha Folding result is that you give it a genetic sequence (biology called the protein primary structure) and run it on a computer to successfully model the protein.

Seemingly simple, but the meaning is extraordinary.

In the past fifty years, techniques for performing the same functions as alpha folding have experimental techniques such as cryo-electron microscopy, nuclear magnetic resonance, or X-ray crystallography. Looking at the instruments alone, they require high prices, but the most important thing is that the cost of the experiment is also very high. It requires a lot of experiments by the experts. The loss of equipment does not mean that the experimental materials cost tens of thousands of dollars.

The main thing is that the whole process is too slow and it takes a few years or even decades for the researcher to spend.

Alpha folding is not the same, biologists no longer need to spend decades in front of the experimental equipment, just need to simply enter the data.

Of course, the benefits of Alpha Folding are not only so, but ultimately it benefits our general public.

蛋白质的新“折叠”!DeepMind重磅推出家族新成员AlphaFold

Taking Alzheimer's disease (known as Alzheimer's disease) as an example, its incubation period in human body is more than ten years, and the cause is complicated. With current medical technology, it is difficult to be clinically difficult. The disease was detected in the year.

In biological research, scientists generally believe that protein changes are the cause of senile dementia. In other words, a certain part of the protein in patients with Alzheimer's disease must be different from normal people. But because our research is too slow, biologists don't know all the protein forms in the body, so it's hard to detect which part is different.

Imagine that it is much easier to detect Alzheimer's disease ten years in advance by Alpha Folding. Just look at the difference in protein length through a computer.

Artificial intelligence's "deep love" of protein structure

This time, it is still a game to make Alpha fold famous.

In 1994, in order to promote research and measure the progress of the latest methods in improving prediction accuracy, the Protein Structure Prediction Technology Key Assessment Community-wide Experiment (CASP) biennial global competition was established. Since its evolution, its results have become the industry standard.

Although there is no such thing as AlphaGo's fame in the past, as a member of the Alpha family, Alpha's folding is not uncommon. In this year's CASP competition, it won the first place with unexpected success.

According to DeepMind's official website, based on deep neural networks, they designed two methods to construct a complete and accurate protein structure.

First, they first collect the angular data between the pairs of amino acids and the chemical bonds connecting these amino acids, and then design these data into analytical tools to assess the structural accuracy of the protein.

Using this analysis tool, the research team came up with the first way to find the best matching protein in the existing protein database. If not, they are based on the closest search structure, replacing it with new gene fragments. To create a new structure that matches the requirements.

And this second method is much simpler. According to them, the researchers mainly use the gradient descent-a mathematics technology, which is higher in accuracy than the first one. Compared to the first method, this technology can predict the entire protein chain in one step, without going through the assembly process, and the whole process is simpler.

DeepMind didn't announce more details, but with this "simple" design, the miracle happened.

Finally, science is coming.

In the human body, protein is a magical existence.

As we all know, protein is the main component of the human body structure, its content is second only to water, accounting for about one-fifth of a person's body weight. Almost all of the functions our body has to perform, including the contraction and stretching of muscles, the perception of light in the body, and the transformation of food, all require proteins to play a key role.

Scientists point out that the structure of a protein largely determines the nature of a protein, so the importance of studying the 3D structure of a protein is obvious. In our bodies, such cases abound, such as the antibody proteins that make up our immune system are "Y-shaped"; collagen is shaped like a rope; CRISPR and Cas9 for gene editing, they are like scissors.

But purely from the gene sequence can only find the three-dimensional shape of the protein is a complex task. According to the traditional research method, scientists need to study from the primary structure and the secondary structure layer by layer, which takes decades or even thousands of years. Only then can the morphological model of the protein be completely established.

The emergence of alpha folding has saved biologists a lot of effort.

On DeepMind's official website, they introduced this major achievement: "We are pleased to share with you the first important milestone in how DeepMind demonstrates how artificial intelligence research can drive and accelerate new scientific discoveries. DeepMind brings together the structure. Experts in the fields of biology, physics, and machine learning have used cutting-edge technology in an interdisciplinary approach to design AlphaFold that predicts the 3D structure of proteins based solely on their gene sequences."

Filter Cartridge

Filter Cartridge


Filter cartridge including : air Filter Cartridge, oil Filter Cartridge, PP Filter Cartridge, string wound Filter Cartridge, stainless steel Filter Cartridge, corrugated Filter Cartridge, spraying Filter Cartridge.

The Filter Cartridge be widely used in vacuum coating, mechanical engineering, CD, metallurgy, petroleum, chemical industry, paper making, food, medicine, electronics, electricity, ceramics, water diversion, printing and dyeing, sugar making .


The advantages of Filter Cartridge :

suitable for filtering strong acid, strong alkali and organic solvents

the area of membrane filtration is large

low diferential pressure, long service life

purifying the air

reduce the noise

save fuel


Filter Cartridge

filter cartridge 1

Filter Cartridge Tube,Water Filter Cartridges,Sediment Filter Cartridge,Iron Water Filter Cartridge

ANPING COUNTY SHANGCHEN WIREMESH PRODUCTS CO.,LTD , https://www.scwpwiremesh.com