Alphafold may be the most exciting scientific innovation of the century. From Google DeepMind First reported In 2020, we will use artificial intelligence to grasp the 3D structure of proteins. This technology has already been used to solve basic biology problems, and was awarded the Nobel Prize (chemistry – to Demis Hassavis and John Jumper) to innovate revolutionary drug discovery. Like most AI, it’s only getting better and just starting out.
The structure of proteins provides clues about their function and helps in designing new drugs. Alphafold, that was it Trained A huge database of experimentally solved structures called Protein Data Bank predicts the structure of a protein based on its amino acid sequence.
In the past, the first step is to use litres of bacteria or viruses to produce huge amounts of protein. you pray To assemble proteins into crystal lattices (notoriously) and then fire high-energy X-rays. This is called X-ray crystallography and can take years. Now Alphafold can do it in minutes (and even a much cheaper hell).
When Alphafold competes for protein structure resolution competition In 2020, it was so good that some people accused the Alphafold team of fraud. With its first appearance, Alphafold has been cutting edge. Currently, the AlphaFold database has approximately 250,000,000 protein structures, which are used by approximately 2 million people from 190 countries. It is used by more people than can perform X-ray crystallography.
He received his PhD in Cancer Biology. I wanted to solve the structure of the protein I worked on. Maybe I could even use it to make new medicines. Now I can go to Alphafold Server And generate the structure in 5 minutes that would consume my entire Ph.D.
Dr. Pauline Lascaux, a molecular and structural biologist at Oxford University, said Alphafold is her heart Latest researchdiscovered a new way for cells to repair DNA, and found that over 90% of the studies she reviews cited it.
So, what does Alphafold drug discovery look like? With this Recent scientific researchresearchers used AlphaFold to predict the structure of serotonin receptors that regulate mood. Through In Silico Testing (on a computer) they tested which of the 1.6 billion (!) molecules could bind to the alphafold structure. What they found was a set of molecules that bind much more closely than drugs produced through a traditional-experimental-approach that could be a new drug for mood disorders.
Alphafold has only been around since 2020, but its impact is meteors. Here are the top three discoveries that have been enabled by AlphaFold so far:
-
solve Problems decades ago: The structure of an internuclear complex, one of the largest structures within a cell. This complex is the guardian of invasion into the nucleus that holds the DNA of the cell. It has to do with cancer, aging, neurodegeneration – and now we know what it looks like on the atomic level.
-
Find a new liver cancer drug. In the laboratory (not patients), drugs target Cancer protein CDK20hindered the growth of liver cancer.
-
Useful for designing “molecular syringes” Therapeutic protein payload into human cells.
There is also a company built in Alphafold. If Alphafold resolves the lock alphaproteo Provide the key. Alphaproteo uses the structure of Alphafold to design molecules that can bind and regulate other proteins. This is used to produce molecules that have never been created before – to target Covid-19, cancer and autoimmunity.
Also, from deepmind, Alpha Misense Tackling the problem of missense mutations – minor changes in genes, uncertain functional effects. Despite their prevalence, we can only tell whether approximately 2% of these changes are pathogenic. Model the structure of the mutation using the AlphaMissense model AlphaFold. If the protein structure changes, it is likely pathogenic. This could transform the diagnosis and treatment of rare genetic diseases.
It is still unclear whether drugs designed using AlphaFold will pass clinical trials successfully, but no one has been tested in humans yet. Only time can be seen.
In the future, Alphafold could discover new drugs by individuals, unlocking unbreakable target drugs, and unlocking molecular secrets. (The other day, Alphafold It helped me to resolve The structure of the sperm oocrop formed during fertilization. )
If the first generation of drug discovery is the natural generation, the biotechnology generation that gave aspirin (from willow bark) and the second gave ozempic, we have now moved to the third generation, the AI generation.
