This article is more than 1 year old
AI-designed drug to treat deadly disease now tested on humans
Software may find solution to terminal idiopathic pulmonary fibrosis
The first phase of clinical trials testing an AI-designed drug to treat a chronic lung disease is now underway, according to biotech startup Insilico Medicine.
Using deep learning to discover new drugs is a relatively fresh development in the pharmaceutical industry, and machine learning algorithms are well-suited to the task. Trained on large amounts of data, they can find patterns in genetic information or chemical databases to come up with novel solutions more quickly that may have been missed by scientists.
But determining whether a new drug designed by automated software will really be effective or not still requires traditional clinical trials. Now, ISM001-055, an "anti-fibrotic small molecule inhibitor," to treat idiopathic pulmonary fibrosis (IPF) will be put to the test with 80 healthy volunteers.
The group will be sorted into ten cohorts; five groups will receive single doses in increasing amounts, while the other five will receive multiple doses in increasing amounts. The goal of this first phase of the clinical trials is to establish the maximum safe dose to start phase two, in which patients with IPF will be enrolled.
Insilico Medicine, based in Hong Kong, uses its Pharma.ai platform to design novel drugs. The system consists of three different main tools. PandaOmics, scours through genetic profiles to discover target biological pathways for which potential drugs can affect. Chemistry42 then generates new structures of molecules and highlights candidates that are stable and possible to synthesize in a real lab. Finally, Inclinico gives a rough prediction of clinical trial outcomes.
"Modern deep learning technologies enable us to perform target identification using longitudinal biological data from healthy subjects and make inferences into a variety of diseases," said Alex Zhavoronkov, founder and CEO of Insilico Medicine, today. "This was the guiding principle for our anti-fibrotic program starting with identifying targets that may play a role in both aging and disease."
- AI tool finds hundreds of genes related to human motor neuron disease
- Free AI protein software packages nearly predicted structure of the Omicron coronavirus variant correctly
- Alphabet launches new AI drug discovery startup, Isomorphic Labs, led by DeepMind CEO
- US drug watchdog green-lights first prostate-cancer-predicting AI software
IPF is a deadly disease. Scar tissue builds up in the lungs over time, making it difficult for patients to breathe. It can lead to chest pain, leg swelling, tiredness, and life-threatening respiratory failure. The cause is unknown, and there are currently no treatments for this terminal illness. ISM001-055 works by preventing scar tissue from forming in the lungs.
"It is a great pleasure to see our AI-discovered novel molecule with a novel mechanism enters the Phase I clinical trial," said Feng Ren, chief science officer of Insilico Medicine.
"We are committed to taking advantage of AI-driven drug discovery to rapidly advance the development of innovative therapeutics for unmet medical needs, especially in the areas of fibrosis, oncology, immunology, and neurology, and initiating this phase I study demonstrates the power of our AI platform in achieving that."
Insilico Medicine said the entire process from target and drug discovery to preclinical candidate testing for ISM001-055 took about 18 months and $2.6m. The Register has asked Insilico Medicine for comment. ®