Generative AI platforms, from ChatGPT to Midjourney, are making headlines in 2023. But GenAI can do more than create image collages and help write emails: it can also design new drugs to treat diseases.
Nowadays, scientists use advanced technology to design new synthetic drug compounds with the right properties and characteristics, which is also known as “de novo drug design.” However, current methods can be labor, time and cost intensive.
Inspired by the popularity of ChatGPT and wondering if this approach could speed up the drug design process, scientists at the Schmid School of Science and Technology at Chapman University in Orange, California, decided to create their own GenAI model, detailed in A new paper, “De Novo Drug Design using Transformer-based Machine Translation and Reinforcement Learning of Adaptive Monte-Carlo Tree Search,” appears in the journal Pharmaceuticals.
Dony Ang, Cyril Rakovski, and Hagop Atamian coded a model to learn a massive data set of known chemicals, how they bind to target proteins, and the rules and syntax of large-scale chemical structure and properties.
The end result can generate countless unique molecular structures that follow essential chemical and biological constraints and effectively bind to their targets, promising to greatly accelerate the process of identifying viable drug candidates for a wide range of diseases, at a fraction of the cost.
DrugAI generates potential drugs never before conceived
In the innovative model, the researchers integrated two cutting-edge AI techniques for the first time in the fields of bioinformatics and cheminformatics: the well-known “transformative encoder-decoder architecture” and “reinforcement learning using Monte Carlo tree search” ( RL-MCTS). The platform, appropriately named “drugAI,” allows users to input a target protein sequence (e.g., a protein typically involved in cancer progression).
DrugAI, trained with data from the comprehensive public database BindingDB, can generate unique molecular structures from scratch and then iteratively refine the candidates, ensuring that the finalists show strong binding affinities with the respective drug targets, crucial for the efficacy of the drugs. potential drugs. The model identifies between 50 and 100 new molecules that are likely to inhibit these particular proteins.
“This approach allows us to generate a potential drug that had never been conceived. It has been tested and validated. Now we are seeing magnificent results”
“This approach allows us to generate a potential drug that has never been conceived before,” Dr. Atamian said. “It’s been tested and validated. Now we’re seeing great results.”
The researchers evaluated the molecules generated by the AI drug according to several criteria and found that the results of the AI drug were of similar quality to those of two other common methods. and in some cases, better. They found that DrugAI’s drug candidates had a 100% validity rate, meaning that none of the generated drugs were present in the training set.
DrugAI’s drug candidates also had drug likeness, or the similarity of a compound’s properties to those of oral drugs, measured, and drug candidates were at least 42% and 75% higher than others Models. Furthermore, all DrugAI-generated molecules showed strong binding affinities to their respective targets, comparable to those identified by traditional virtual screening approaches.
Ang, Rakovski and Atamian also wanted to see how DrugAI’s results for a specific disease compare to known existing drugs for that disease. In a different experiment, screening methods generated a list of natural products that inhibited COVID-19 proteins; drugAI generated a list of new drugs targeting the same protein to compare their characteristics. They compared drug likeness and binding affinity between natural molecules and AI drugs, and found similar measurements in both, but AI drugs were able to identify them in a much faster and less expensive way.
Additionally, the scientists designed the algorithm to have a flexible structure that allows future researchers to add new features. “That means we’ll end up with more refined drug candidates with an even higher chance of ending up being a real drug,” Dr. Atamian said. “We are excited about the possibilities that exist in the future.”
