“The integration of AI, robotics and ageing research will allow us to find complete cures to very complicated diseases like Alzheimer’s and Parkinson’s, and many others,” Zhavoronkov said, adding that AI has the potential to provide humans with tools to completely avoid these diseases.
In 2014, the company started training deep neural networks to understand human ageing, leveraging AI capabilities to record, track and analyse people’s health throughout their lives.
“AI can understand billions of people just by understanding ageing. It can then start understanding the basic biology of diseases, and not only to slow them down,” Zhavoronkov said.
“In the ideal scenario, you want to ensure that the disease completely disappears or it does not happen at all.”
Zhavoronkov, who refers to ageing as “biology in time”, is an expert in generative biology and chemistry, and in the research of ageing and longevity.
“Human biology and the homeostasis (state of balance among body systems) of your body degrades over time. That is what is happening in a disease. Diseases accelerate this process, or are caused by this process. So without understanding the process of basic human ageing, you will not understand most of the diseases,” he said.
Idiopathic pulmonary fibrosis (IPF) results in chronic scarring of lung tissue that makes breathing difficult. The disease affects 5 million people worldwide, mostly above the age of 60, and has a high mortality rate. The median survival rate of untreated patients is two to three years.
There is no known cause of the disease and no cure is available, but some treatments can help alleviate symptoms and slow its progression. Many patients who receive steroids suffer from progressive decline in lung function and succumb to respiratory failure.
Chinese researchers hope to create ‘real AI scientists’
Chinese researchers hope to create ‘real AI scientists’
In the new study, scientists used generative AI to find an anti-fibrotic target and its inhibitor, significantly shortening the traditional drug development timeline which can often span more than a decade.
“This work was completed in roughly 18 months from target discovery to preclinical candidate nomination and shows the capabilities of our generative AI-driven drug-discovery pipeline,” the team said in an article published in the peer-reviewed journal Nature Biotechnology on Friday.
The researchers first trained a target identification engine in Insilico’s AI platform using data and publications about fibrosis. The condition results in a thickening or scarring of tissues that can reduce organ elasticity.
Fibrosis is closely related to the process of ageing, which generates chronic inflammation resulting in fibrosis.
02:51
Biological robots made from human cells show promise for use in medical treatment
Biological robots made from human cells show promise for use in medical treatment
With the help of a predictive AI approach, a protein abbreviated as TNIK emerged as the top anti-fibrotic target. The team then used a generative chemistry engine to generate about 80 small-molecule candidates to find the optimal inhibitor, known as INS018_055.
“[The inhibitor] exhibits desirable drug-like properties and anti-fibrotic activity across different organs … through oral, inhaled or topical administration,” the scientists wrote.
The study “provides evidence that generative AI platforms offer time-efficient solutions for generating target-specific drugs with potent anti-fibrotic activity”, they said.
“We believe that this study underscores the strength of AI-enabled drug-discovery approaches, which will likely revolutionise drug discovery.”
‘Extra pair of eyes’: Hong Kong researchers find AI helps doctors spot tumours
‘Extra pair of eyes’: Hong Kong researchers find AI helps doctors spot tumours
When asked about the importance of artificial intelligence in the drug discovery research, Insilico’s AI chatbot, which is based on ChatGPT, said: “By streamlining the initial stages of drug discovery, AI enables us to reach the clinical trial phase more rapidly, focusing resources and efforts on these critical stages of development.
“While AI has the potential to accelerate early-stage drug discovery tasks such as target identification and lead optimisation, it does not substantially reduce the duration of clinical trials.
“Clinical trial phases still require extensive time for ethical and regulatory approval, patient recruitment, treatment duration and data analysis.”
Source: scmp.com
