Proteins are the building blocks of life, and defining their activities is a necessary step in understanding a wide range of life processes. The most complex proteins, on the other hand, are formed only from more than 20 amino acids arranged and combined in different ways, which are swiftly folded into a specific three-dimensional structure within milliseconds to several milliseconds. The various structures that protein structures take account for their diverse functions.
Protein structure prediction is a fascinating field in the life sciences that has attracted many researchers to tackle difficult issues, but it has always faced difficulties, high costs, and sluggish progress until the advent of artificial intelligence, single protein folding. This century’s problem has been mostly solved thanks to this revolutionary development. The world has entered a period of exponential growth in life science and technology as a result of this revolutionary discovery.
The new crown pneumonia epidemic has fueled AI + synthetic biology growth, and AI protein design is now becoming. Tianyuan, a Chinese artificial intelligence firm based on the self-developed AI-led protein design platform TRDesign, just developed a brand-new coronavirus spike protein (S protein) binding agent from scratch that can prevent the spike protein from attaching to human cell ACE2 receptors.
Tiantian’s design method, which involves the use of mini-proteins, has the advantages of low molecular weight, high thermal stability and low cost. In comparison to traditional methods for manufacturing secondary spike protein conjugates that employ antibody or natural ACE2 transformation, Tiantian’s technique has the benefits of smaller size, enhanced thermal stability and lower cost of mini-proteins; this calculation approach is more efficient and provides greater folding space. It’s thick and may effectively shield the entire interface between ACE2 and the new coronavirus spike protein from a variety of viral mutations, which is the first in China.
Design process of Tianrang new crown spike protein binder
Protein design with AI at its core has gone from a technological concept to a value verification stage, and several AI-led biological enzymes and pharmaceutical proteins have progressed through the clinical study phase. The strong entry of domestic businesses has hastened technological progress in this field, laying the groundwork for further strengthening synthetic biotechnology’s strategic scientific and technological strength. It is anticipated that a large number of institutions and enterprises will enter the waters of such technological breakthroughs in the next few years.
Spring in Synthetic Biology
Synthetic biology is becoming a disruptive force in the bioeconomy, gradually transitioning from theoretical study to a future manufacturing system. Synthetic biology, in its broadest sense, refers to the creation and modification of cells or living organisms by constructing biological functional components (such as enzymes), devices (e.g., sensors), and systems that have biological functions that match human requirements.
Synthetic biology technology can create almost everything that humans require, from fragrances and textiles to food and energy, because it can be used to make many of the things produced by today’s industry.
Synthetic biology is expected to liberate the supply chain from constraints imposed by material availability. Companies may start from the ground up and build anything they want with cells because there are no limitations to raw material supply. Bovine muscle cells, for example, may produce almost 4.4 billion pounds of beef, more than Mexico consumes in a year. Science-based start-ups have emerged as a result of synthetic biology; they’re attempting to convert outdated products and processes.
In the United States, synthetic biology’s development is still in its early stages. In recent years, several technology roadmaps have been published in quick succession. Synthetic biology is one of the key competitive technologies under consideration in the “2021 American Innovation and Competition Act.”
Synthetic biology, artificial protein, and other industries are also developing rapidly in China. At the same time, China is speeding down the synthetic biologist highway. The “14th Five-Year Plan for Bioeconomic Development,” released by the National Development and Reform Commission on May 10, 2015, includes a section on synthetic biology. It is obvious that it is critical to speed up the advancement of syntheticbiology, artificial protein technology, and other biotechnology sectors as well as promote deep integration between biotech and IT technologies.
Start-ups in the synthetic biology sector are exploding, and investors and venture capital are flocking to the market. In 2020, global firms working in synthetic biology will have received US$7.8 billion in funding and investment, nearly twice as much as last year. Synthetic biology start-ups raised $6.1 billion in investment and financing during the third quarter of 2021, a rise of 33%. Health and medical research investments grew by far more than other sectors.’
According to the Boston Consulting Group in February 2022, it is expected that by the end of this century, synthetic biology will be widely used in manufacturing, which accounts for more than one-third of global output, creating a value of 30 trillion US dollars. Some industries may be more vulnerable due to developments in real-time data collection, automation and artificial intelligence.
In the following five years, synthetic biology industries such as healthcare and cosmetics, medical devices, and electronics will face rivals in this field. The pharmaceutical and food industries already confront competition.
In recent years, deep learning AI algorithms have grown in power and efficiency, with their exceptional skills in continuous learning of huge data and innovative exploration of unknown territory effectively matching the needs of today’s synthetic biology engineering trial-and-error platform. Information mining from high-dimensional correlation information has a lot of promise.
AI-led de novo design can be used to create specific biological elements that serve a certain function. This can replace some of the links that need to be experimentally tested, saving time and money. Additionally, many proteins that were once inaccessible can now be studied directly using design, which will accelerate their application in synthetic biology. Finally, learning engineering is being implemented to improve AI’s ability to design biological elements.
Since 2020, when the capital market and technology had major breakthroughs, more than 10 companies from all around the world were successfully listed. More than 30 candidate compounds involving AI technology also entered the clinical stage during that year. Recently, China’s AI-centered drug design industry has developed rapidly. This is mainly due to advances in molecule generation, activity prediction, and virtual screening techniques. Most of these improvements aim to increase the efficiency of drug research and development. However, a few AI-led de novo design companies have the potential to make great strides in terms of technological advancements. These companies could eventually change the fundamental research and development method for drugs altogether.
Protein design opens new valve
The living system is extremely complex, with a large number of genes and regulatory components, and the components build modules and networks with many different combinations that are difficult to accurately convey and anticipate. In the human body alone, trillions of cells engage and provide feedback. By contrast, the process of obtaining cells and their interactions is challenging, requiring time-consuming trial and error correction.
Protein structure design is a cutting-edge life science technology. Protein structure analysis has several difficulties, including “high technical requirements, high equipment cost, and lengthy experiment time,” which make obtaining structural information difficult. This severely hampers the study and interpretation of sequence-structure-function in protein engineering due to these issues. R&D The use of AI will make the acquisition of high-throughput and precise protein structure information a reality, allowing for significant acceleration in protein engineering design.
Protein engineering may be used to create novel protein drug prospects much more quickly than has been possible previously by using natural proteins with known functions in humans, which are little subsets of the enormous protein space. New protein drug candidates can be rapidly developed for previously untreatable diseases through protein design, lowering the risk of drug development and fundamentally altering how drugs are developed.
Protein structure prediction and design are both disciplines that are concerned with the fundamental issue of protein folding, and they are sometimes seen as reciprocating hypotheses. The Generative Biology platform, which is drove by machine learning, underpins Generate Biomedicines, which was incubated by Flagship Pioneering, a top VC in the life sciences.
It has dissected millions of proteins and learnt their functional codes in order to swiftly develop and manufacture new protein therapeutics that target specific goals. And, in January 2022, it was given 5 orders from Amgen, with a pre-payment of $50 million and the possibility for a contract worth more than $1.9 billion.
Previously, TRFold, a domestic inventive firm specializing in general intelligence research, developed its own protein structure prediction platform TRFold that is comparable to AlphaFfold2 and produced an end-to-end de novo protein design, detection, stability, and affinity optimization.
Protein design may be complex and difficult. It requires continual iterative adjustments of the main and side chains, according to AI-based technology. Protein folding reverse mapping and rapid protein synthesis based on the target function are all possible as a result of this research.
In addition to Tianliang, there have been several domestic firms that have addressed important issues in the field, such as Zhiyu Biotechnology and Molecular Heart. Zhiyu Biotechnology is capable of transforming and designing pharmaceutical intermediates’ enzymes from natural substrates using pure AI algorithms and design techniques in a short period of time. The data demonstrate that the modified enzyme’s catalytic activity was increased by 5-7 times, while its thermal stability improved by about 30 degrees.
The cost of developing a new medication, according to “Nature,” is about $2.6 billion USD, which takes around 10 years. Phase III clinical trials, as well as registration approval, are included in the process. However, less than 10% of those that can pass these tests and are successfully marketed do so.
Xue Guirong, the founder and CEO of Tianyang, says that their AI-led protein design method is different from traditional methods. Their platform is based on AI and in the future they will also have a wet laboratory that will use AI to design proteins. They believe that AI has great potential to understand life and design new proteins.
“Protein drug design is the crown jewel of biopharmaceuticals,” says Baidu Ventures CEO Gao Xue. AI’s excellent success in protein structure prediction enables it to create AI-based protein medicines from the ground up.
The Molecular Heart team has conducted research and commercial applications in the field of AI protein prediction and design, as well as developed an AI macromolecule optimization and design platform “MoleculeOS” with independently owned intellectual property rights using data-driven deep learning methods to assist biotechnology professionals in rapidly identifying and generating the most suitable proteins.
According to Ms. Li, a partner at Sequoia China, using AI software to predict protein structures may lead to the development of new proteins that bind to certain locations based on the target structure. This breakthrough is expected to have a big impact on drug discovery.
The development of recombinant protein drugs will cause a boom in the future. This is because of the continuous escalation of new coronavirus mutations. If AI’s de novo design is integrated with drug development processes, it will be possible to produce large quantities of high-value protein drugs.
