Drug Discovery with Generative AI: The Insilico Approach
With the application of generative AI, the company is now poised for a revolution in drug discovery that will not only speed up the development of cancer medications but also result in significant financial savings.
According to Alex Aliper, co-founder and president of Insilico Medicine, “the entire [drug discovery] process is extremely inefficient, slow, and expensive so it takes over 10 years to develop a drug.” Alex Aliper spoke exclusively with Arabian Business.
Currently, bringing a medicine to market takes between 10 and 12 years and costs close to $2 billion. This is a significant problem that Big Pharma has been battling for many years.
Pharma.AI, an end-to-end drug discovery platform from Insilico, streamlines the procedure by pinpointing illness targets, producing data on novel compounds, and forecasting the results of clinical trials. PandaOmics, Chemistry42, and inClinico are the three parts of the AI-powered platform that carry out the drug discovery process using millions of data samples and various data kinds.
Aliper thinks that the drug discovery engine can find medications “as efficiently and inexpensively as possible.” It is the first project of its kind in the UAE that is currently being managed from Abu Dhabi: the creation of the Pharma.AI platform. Given that no drug has ever been found in the Middle East, this achievement seems implausible. The company’s creator and CEO, Dr. Alex Zhavoronkov, called in with Aliper from Chengdu, China, saying he “hopes to change that.” Currently, there are approximately 10,000 medical requirements in the globe that have no available treatments. Due to this, more improved drug discovery is urgently needed.
– And what better way to accomplish this than with generative AI?
We can see that conventional approaches in Pharma are failing, therefore Zhavoronkov claimed that “only with the power of AI are we able to kind of crack this problem.”
The relevant target for the disease of interest is found using Insilico’s AI platform, which can locate a protein or other entity within an organism. After that, the target is duplicated to modulate the illness and either “reverse it or ameliorate it,” according to him.
The platform includes three products: Chemistry42, which finds novel compounds with modulator targets found in PandaOmics; PandaOmics, which covers target identification – the first step in the discovery process; and inClinico, a tool created especially to forecast the probability of success in a phase 2 clinical trial.
Phase 2 of the drug development process is the most important stage because the success rate is frequently fairly low.
“Approximately 66% of trials fail in phase 2, which evaluates a drug’s effectiveness in humans. So, to increase the efficiency of the development process and save a ton of money on resources and time, we employ inClinco to forecast which medications would be successful in passing phase 2, according to Zhavoronkov.
According to Zhavronkov, who is also a highly recognized expert in the fields of generative chemistry and biology, the US Food and Drug Administration (FDA) only approved 50 medications last year. Among those 50 medications, only seven are “more or less innovative” compared to the others, which are only modified medications, he continued.
No medications have yet been found in the Middle East that have been approved in the US since drug discovery is such a time-consuming, expensive, and high-risk procedure. Drugs from several Big Pharma companies hardly ever advance to phase 1 clinical trials. Since most of these efforts began in 2019 and 2020, the FDA has never approved any AI-designed medications, although there were some early attempts.
