Technology

AI Develops Drug for Treating Pneumonia Inflammation

AI Develops Drug for Treating Pneumonia Inflammation

Artificial intelligence can bring enormous operational benefits to humanity, as it will be possible to develop drugs quickly with its assistance and at lower costs. Insilico Medicine, one of the largest international biotechnology companies, announced that a drug discovered by artificial intelligence has successfully passed the world's first phase of clinical trials. Now, AI is not only capable of creating images, writing texts, and composing music but is also actively contributing to medicine.

The drug, named ISM001-055, is being developed to treat idiopathic pulmonary fibrosis (IPF). The causes of this disease are not entirely clear, hence the term "idiopathic." This disease primarily affects the elderly and is very troubling, as lung tissues become scarred and stiff, potentially leading to the patient's death. Each year, 5 million people worldwide are diagnosed with this disease, and the average life expectancy after diagnosis is 3 to 4 years. The drug is expected to serve as a "compensation" for them; it not only aims to help the patient's lungs but is also intended to be a tool for rejuvenating the vitality of the human body as a whole.

In fact, AI did not work alone but collaborated within a team. In the initial phase of the drug development project, called PandaOmics, AI utilized all medical data on the subject in search of a protein that could cause pulmonary fibrosis (the excessive growth of connective tissue leading to scarring). Once the target protein was identified, it was connected to another AI called Chemistry42, which is known as a generative adversarial network (GAN). Its task was to create a molecule that could become an effective treatment for the disease. The GAN considers different directions, generating part of its mind's suggestions while another part critiques them. Thus, this "two-in-one" approach means "I come up with a solution myself, and I also look at its weaknesses." As a result, the most suitable option is determined. Once the AI created the molecule, a small batch of the potential drug was quickly manufactured in the laboratory.

In the third phase, AI monitored how that substance worked on 8 volunteers in Australia. The entire drug development process took about 2.5 years and cost £2.6 million, which is a drop in the ocean compared to the classical method of drug production. All eight brave Australian volunteers felt satisfied after the first course of taking the drug, so the developers decided to move forward.

Our readers are reading too