The use of large language models and GenAI applications like ChatGPT has gone past the initial functionalities, like text writing and researching. Not that those particular use cases are any less helpful, but scientists are beginning to deploy the same GenAI models to solve more complex problems.
Scientists Create AI Model to Speed up Drug Discovery Process
ScienceDaily reported Wednesday that a group of scientists in the Schmid College of Science and Technology at Chapman University in Orange, California, have developed their own ChatGPT-like model to accelerate the processes involved in the drug design.
The traditional drug design process is notoriously slow, expensive, and involves a lot of trial and error. However, the scientists, namely Dony Ang, Cyril Rakovski, and Hagop Atamian, are able to identify viable drug candidates for several diseases much faster and at a better cost using the model called DrugAI.
The scientists created the model by combining two cutting-edge AI techniques (the “Encoder-Decoder Transformer architecture” and “Reinforcement Learning via Monte Carlo Tree Search” (RL-MCTS)) for the first time in the fields of bioinformatics and cheminformatics.
The model was reportedly trained on a vast public database of chemicals to learn how they target proteins involved in cancer progression and other diseases and every other property crucial for the efficacy of potential drugs.
Per the report, DrugAI was able to identify 50-100 new molecules likely to inhibit disease-causing proteins.
“This approach allows us to generate a potential drug that has never been conceived of,” one of the researchers said. “It’s been tested and validated. Now, we’re seeing magnificent results.”
Is AI the Future of Medicine
The AI model’s candidate drugs were found to have a 100% validity rate. They were also reportedly measured for drug-likeness, and DrugAI candidate drugs were at least 42% and 75% higher compared to other models.
Given the progress, it’s safe to say that the potential of AI technology in scientific advancement is undeniably significant. Especially in the field of drug discovery, the application of technology holds immense promise for finding potential treatments for life-stealing diseases.
Source: https://www.cryptopolitan.com/new-chatgpt-like-model-identifies-drug-candidates/