New documentary from Insilico Medicine reveals the story of its lead AI-designed drug

A new documentary from clinical stage artificial intelligence (AI)-driven drug discovery company Insilico Medicine (“Insiico”) takes viewers behind the scenes of the company’s lead drug, INS018_055, which was discovered and designed using the company’s generative AI platform. The documentary is a companion to a related paper in Nature Biotechnology, “A small-molecule TNIK inhibitor targets fibrosis in preclinical and clinical models,” which provides in-depth insights and extensive data around this lead program – a potentially first-in-class small molecule drug designed to treat the devastating lung condition idiopathic pulmonary fibrosis (IPF), a disease marked by excessive scarring of the lungs that affects 5 million people worldwide and has few treatment options. The lead drug represents a milestone achievement in AI drug discovery – the first generative AI-designed drug for an AI-discovered target to advance all the way to Phase II trials with patients. 

In the documentary, Insilico Medicine founder and CEO Alex Zhavoronkov, PhD, describes how Insilico’s AI drug discovery differs from traditional methods. “In a traditional model,” he says, “you start with a bunch of targets and you have lots of shots on goal. The probability of identifying the right target in a disease at the preclinical level is about 1% to 5% and it can cost you billions of dollars. So we started training deep neural networks on human age using multiple data types like gene expression and protein expression. By doing that, you train deep neural networks to learn basic human biology at every level.” 

The company first proposed the concept of using generative AI for the design of novel molecules in a peer-reviewed journal in 2016. After further development and validation, that evolved into the company’s generative adversarial network (GAN)-based AI platform which was ultimately integrated into the commercially available Pharma.AI platform, which includes generative biology, chemistry, and medicine. This platform not only produced Insilico’s lead drug, but with 30 others currently in development in cancer, fibrosis, IBD, and COVID-19, with 6 in clinical stages. 

“We started employing this technology for making deep neural networks imagine new molecules just like a human would imagine,” Zhavoronkov says in the documentary. “I can say ‘I want to see a molecule with this number of hydrogen bonds in the right places and I want this molecule to fit in this pocket but also be metabolically stable and drug-like.’ AI is ‘imagining’ new molecules with the desired properties – and it is imagining many.” 

Insilico’s AI platform ultimately selected TNIK as a promising target for a new IPF treatment. IPF is an aggressive, interstitial lung disease and current treatments aimed at inhibiting other pathways have been unable to slow, stop, or reverse progression of the disease. 

As the documentary describes, “to identify novel kinase targets, Insilico scientists restricted the small molecule filter to the highest level, and the safety and novelty filter to the medium level. Only targets that belonged to protein kinase or receptor kinase and with available structures were included in the calculation.” Scientists used five OMICs score models – mutated sub modules, causal inference, epression, heterogenous graph walk, and matrix factorization. TNIK emerged as the number two target.

Next, Insilico’s expert scientists stepped in, as Feng Ren, PhD, co-CEO and Chief Scientific Officer, describes in the documentary. “We used our internal filters for target validation and we nailed down TNIK as the target we wanted to pursue,” he says. “Once we gave Chemistry42 [Insilico’s generative AI drug design engine] the structure of that protein, it gave us hundreds of compounds.” 

Insilico’s medicinal chemists selected 78 of those compounds for further testing. Molecule 055 stood out as the most promising – with a remarkable binding ability and inhibition of the target. And when tested in animal models, the molecule improved lung function, and diminished fibrosis and inflammation. “We got a really great result in the animal model,” Ren said. “That was a turning point.”  

INS018_055 was nominated as a preclinical candidate – a record 18 months from target discovery — and quickly accelerated into further trials, from first-in-human, to Phase I trials with healthy volunteers, where the drug showed excellent results, delivering a favorable pharmacokinetic and safety profile. 

Now, the drug is currently in Phase 2a trials in both the U.S. and China with IPF patients – less than 4 years from discovery. As noted in the film, Insilico is pursuing an inhalable formulation of INS018_055 as well, and exploring its potential as a treatment for kidney fibrosis. 

Now, says Insilico Medicine President Alex Aliper, PhD, in the documentary, “We want to perfect the platform and enable others to use the platform to accelerate the drug discovery and development cycle and generate more novel therapeutics for patients in need.” 

About Insilico Medicine 

Insilico Medicine, a global clinical-stage biotechnology company powered by generative AI, connects biology, chemistry, and clinical trial analysis using next-generation AI systems. The company has developed AI platforms that utilize deep generative models, reinforcement learning, transformers, and other modern machine learning techniques for novel target discovery and generating novel molecular structures with desired properties. Insilico Medicine is developing breakthrough solutions to discover and develop innovative drugs for cancer, fibrosis, immunity, central nervous system diseases, infectious diseases, autoimmune diseases, and aging-related diseases. www.insilico.com