3 outstanding AI trends in drug discovery in 2022

Without a doubt, the year 2022 has seen a frenetic ride of artificial intelligence innovation and business use cases in many industries. artificial intelligence has expanded beyond marketing, customer satisfaction and employee retention. One area where it has made great strides is medicine, biotechnology, and pharmacology, where it is transforming Research and make medicine and develop.

The Price discovery and development of a drug is on average $1.3 billion and “requires anything from 12 to 15 years to hit the market,” according to one PubMed paper. So it’s no surprise that the drug discovery industry has seen a significant rise in AI-powered technologies. A case in point is an article in Nature that noted that integration of AI into drug discovery and development workflows has grown by nearly 40% year-over-year.

According to healthcare investors Tzvi Bessler and Morris Laster, Ph.D., “drug discovery companies are taking advantage of AI in a variety of ways, such as using machine learning algorithms to identify drugs potential drug candidates, predict their efficacy and safety, and optimize their design. . For example, they use AI to analyze large data sets of biological and chemical information to identify patterns and relationships that may be relevant to drug discovery.”

This, they say, helps companies “identify promising leads and accelerate drug discovery.”

As the year of AI comes to a close, VentureBeat spoke with some of the experts about the hottest AI trends of 2022 in drug discovery. Here are three prominent trends:

1. More effective in biological modeling and drug target discovery

James Handler, a professor at Rensselaer Polytechnic Institute and president of the Computer Technology Policy Council Consortium, told VentureBeat about two uses where AI is showing great promise in drug discovery: reducing drug discovery. number of potential candidates for trials and offer a potential explanation for the drug’s secondary use — that’s why a drug shown to be effective for a condition it was not originally designed for plan for treatment.

In both cases, he noted, it is important that “AI can reduce the number of possibilities that need to be explored through traditional means.” This aids in biological modeling and drug target discovery. “However,” he added, “an important aspect of this is that AI systems can explain their predictions to humans, a focus of the present. search. This allows humans to make the final decision [on] analysis and testing, with AI greatly reducing the cost of bringing successful drugs to market.”

Drug discovery and development often begins with the identification of a biological target – such as a gene, protein, receptor, or enzyme. Protein is most Common drug target because of its ability to influence cell behavior or function. Thus, traditional drug discovery efforts involve the selection of specific proteins with vesicles that can be influenced by promising drug-like molecules (which then become ligands or bound drugs). ).

However, this process is computationally challenging. Of the 20,360 human proteins stored in SWITZERLAND-PROT — the world’s most widely used, professionally curated protein sequence database — only a few have been discovered as drug targets.

Organizations are now using AI’s ability to correlate and match large amounts of data, leading to more effective drug target identification and discovery. By 2022, many AI-powered healthcare businesses have shifted resources to building advanced modeling tools that not only model biology, but also identify and validate new targets. This year, major pharmaceutical enterprises such as AstraZeneca and Pfizer have partnered with AI providers to provide targeted discovery services for discover more than eight new goals.

2. Improved protein structure prediction

Proteins need to be folded into specific three-dimensional structures. Incorrect folding or absenteeism has been link to the pathology of many diseases. Protein structure prediction is also relevant in drug discovery because it helps to better understand how a protein works, which in turn indicates how it can be influenced, controlled, and modified.

This is a difficult task, however. A computational biology Research report notes that predicting protein structure “remains a common challenge.”

However, 2022 has inspired significant progress in predicting how proteins fold. This is led by DeepMind innovation open source software, AlphaFold, it is possible to predict the 3D structure of a protein from its one-dimensional amino acid sequence. AlphaFold was able to predict the protein structure of “nearly all the cataloged proteins known to science”.

Reducing the time that would normally take years to just a few seconds, in July the software used the power of deep learning AI to predict and publicly share more than 200 million protein structures belonging to animals, plants, bacteria, fungi and other organisms.

In November, DeepMind’s AI model found a worthy rival in Meta’s research team. Meta leverages AI’s natural language processing (NLP) capabilities and applies “big language models” to predict the structure of more than 600 million proteins found in both known and unknown organisms. This is a major advance for protein structure prediction, which was previously a major challenge.

During new drug design (DNDD) — which PubMed describe is “the design of new chemical entities that conform to a set of constraints using computational growth algorithms” — molecules developed from the ground up, allowing for short trial and error periods than. As de novo is usually a reproduction type of design, it is mainly based on computational processes and study carefully model.

The year 2022 has seen significant progress in the development of combined de novo approaches consolidate-Learn architecture in regular AI neural network.

Virtual sifting of existing databases, another aspect of drug design, is also an object of attention in 2022. Scrutinizing large databases for similarities and uncovers specific peculiarities that are defining AI features. The pharmaceutical giants have applied this technology to large volumes of databases and invested millions of dollars in partnerships with AI platforms capable of virtual screening of trillions of synthetic compounds.

Handler noted that drugs that appear to be effective in animal testing often fail when they are tested in humans. The challenge, he said, is to predict toxicity from previous data. “New techniques are exploring how to use AI models that integrate a variety of test data to better predict toxicity and thus reduce the number of candidates requiring costly testing.”

Handler added that more data is available and shared, and guess that “this will create many innovation opportunities in drug discovery” by 2023. As VentureBeat reporter Ashleigh Hollowell noted in a recent article article“Progress, not perfection, is what to expect [from AI applications] in 2023” — even in the complex world of drug discovery.