New AI-based method can lead to faster, low-cost development of protein-based drugs

Thanks to a newly published study by scientists from the Chalmers University of Technology, Sweden, artificial Intelligence (AI) can now create novel and functionally active proteins.

New AI-based method can lead to faster, low-cost development of protein-based drugs
Researchers at Chalmers University of Technology, Sweden, present a way to generate synthetic proteins using Artificial Intelligence. The new approach has huge potential for developing efficient industrial enzymes as well as new protein-based medicine, such as antibodies and vaccines. Image: Pixabay/Yen Strandqvist, Chalmers University of Technology. Image Credit: Pixabay/Yen Strandqvist, Chalmers University of Technology.

What we are now able to demonstrate offers fantastic potential for a number of future applications, such as faster and more cost-efficient development of protein-based drugs.”

Aleksej Zelezniak, Associate Professor, Department of Biology and Biological Engineering, Chalmers University of Technology

Proteins are essentially huge, complex molecules that play a major role in all living cells by constructing, altering, and breaking down other molecules naturally within the cells. Proteins are also extensively employed in industrial products and processes, and in everyday lives.

For example, protein-based medications are quite common—insulin is one of the most commonly prescribed diabetes drugs. A few of the most effective and costly cancer drugs are also based on proteins, including the antibody formulas presently being used for treating the COVID-19 disease.

From computer design to working proteins in just a few weeks

Existing techniques used for protein engineering depend on introducing arbitrary mutations to protein sequences. But whenever an extra random mutation is introduced, the activity of proteins reduces.

Consequently, one must perform multiple rounds of very expensive and time-consuming experiments, screening millions of variants, to engineer proteins and enzymes that end up being significantly different from those found in nature.”

Aleksej Zelezniak, Associate Professor, Department of Biology and Biological Engineering, Chalmers University of Technology

Zelezniak, who is also a research leader, continued, “This engineering process is very slow, but now we have an AI-based method where we can go from computer design to working protein in just a few weeks.”

The latest findings from the Chalmers team were recently reported in the Nature Machine Intelligence journal and signify a major innovation in the domain of synthetic proteins. Zelezniak’s research team and colleagues have designed an AI-based method, known as ProteinGAN, which makes use of a generative deep learning strategy.

The AI is basically given a huge amount of data from well-researched proteins; it analyzes this data and tries to produce new proteins based on it. Simultaneously, another portion of the AI attempts to determine whether the synthetic proteins are real or not. The proteins are sent to and fro in the system until the AI can no longer distinguish between synthetic and natural proteins.

This approach is widely known for producing videos and photos of individuals who do not exist; however, in this analysis, the technique was used for generating highly different variants of proteins with naturalistic-like physical traits that could be analyzed for their functions.

Extensively used in daily products, the proteins are not always completely natural but are made using protein engineering and synthetic biology approaches. With the help of these techniques, the original sequences of proteins are altered in the hope of producing synthetic new protein variants that are more stable, efficient, and customized toward specific applications.

The latest AI-based technique is crucial for designing efficient industrial enzymes and even new protein-based therapies, like vaccines and antibodies.

A cost-efficient and sustainable model

Martin Engqvist, an Assistant Professor also from the Department of Biology and Biological Engineering, was involved in developing the experiments to validate the proteins produced by the AI-based technique.

Accelerating the rate at which we engineer proteins is very important for driving down development costs for enzyme catalysts. This is the key for realizing environmentally sustainable industrial processes and consumer products, and our AI model, as well as future models, will enable that. Our work is a vital contribution in that context.”

Martin Engqvist, Assistant Professor, Department of Biology and Biological Engineering, Chalmers University of Technology

Zelezniak added, “This kind of work is only possible in the type of multidisciplinary environment that exists at our Division – at the interface of computer science and biology. We have perfect conditions to experimentally test the properties of these AI-designed proteins.”

The following step for the team is to investigate how the AI technology could be utilized to particularly improve the protein properties, like increased stability, something which could provide an excellent advantage to proteins employed in industrial technology.

Journal reference:

Repecka, D., et al. (2021) Expanding functional protein sequence spaces using generative adversarial networks. Nature Machine Intelligence.


The opinions expressed here are the views of the writer and do not necessarily reflect the views and opinions of AZoLifeSciences.
Post a new comment
You might also like...
Structure of amyloid fibers formed by the protein hnRNPDL-2 determined