Insilico Medicine announces preclinical drug discovery benchmarks from 2021 to 2024

Insilico Medicine ( "Insilico") , a clinical stage generative artificial intelligence (AI)-driven biotechnology company today announced a set of preclinical drug discovery benchmarks from the 22 developmental candidate nominations achieved by its platform from 2021 to 2024. These benchmarks underscore the platform's efficiency and represent a potential new standard for the drug discovery industry by significantly reducing developmental times, cost, and by allowing resources to be redirected toward further research and development.

The promise of AI-powered drug discovery (AIDD) has always revolved around three factors: speed, cost, and the probability of success. Tens of billions of dollars have been invested in AIDD since the dawn of the deep learning revolution, which started with notable achievements where deep neural networks (DNNs) came close to or exceeded human capabilities in the ImageNet competition, "Atari Game", and other notable demonstrations. 

Insilico Medicine started in 2014, and until 2019 bootstrapped solving complex problems across end-to-end drug discovery and development for big pharmaceutical, biotechnology, and consumer companies, prioritizing projects focusing on longitudinal data. In 2019, Insilico published its Generative Tensorial Reinforcement Learning (GENTRL) concept in Nature Biotechnology with experimental validation setting a benchmark of 46-days from project initiation to animal pharmacokinetic (PK) studies, including synthesis and multiple preclinical assays. These demonstrations were followed by Insilico's first large $37 million Series B financing round announced in September 2019, which launched the AI software business and started Insilico's internal drug discovery program. In February 2021, Insilico Medicine nominated its first developmental candidate in lung fibrosis, setting a benchmark of 18 months from project initiation to DC nomination. 

By December 31, 2024, Insilico Medicine nominated 22 developmental candidates, with 10 programs progressing to human clinical stage, completed 4 Phase I clinical studies, and completed 1 Phase IIa study in idiopathic pulmonary fibrosis (IPF) - demonstrating good safety and dose-dependent efficacy. 

In this release, Insilico Medicine announces its internal benchmarks for DC nomination timelines as well as how these developmental candidates are defined. 

The terminology and standards for preclinical milestones vary from company to company. Insilico defines the typical developmental candidate (DC) as the stage after which only IND-enabling study remains before the drug enters human clinical trials. The typical DC package at Insilico Medicine specifically includes, but is not limited to, the following components:

• Enzymatic assays demonstrating binding affinity

• In vitro ADME profile

• Microsomal stability assays

• Mouse/rat/dog pharmacokinetic (PK) studies

• Cellular functional assays and PD marker validation demonstrating target engagement

• In vivo efficacy studies and PK/PD/efficacy analysis demonstrating target engagement and identifying efficacious dose ranges

• non-GLP toxicity studies across multiple species

Benchmarks: 

Insilico has now officially announced key timeline benchmarks for its 22 developmental candidates, emphasizing its commitment to efficiency, transparency and innovation in drug development:

• # of Developmental Candidate Nominations (2021–2024): 22 candidates

• Longest time to DC: 18 months; 79 molecules synthesized 

• Average time to DC: ~13 months; ~70 molecules synthesized per program

• Shortest time to DC: 9 months

  • QPCTL program co-developed with Fosun Pharma and in ongoing Phase I trial

The success rate for advancing programs from the developmental candidate stage to the IND-enabling stage has been 100%, excluding programs that were voluntarily discontinued by the company for strategic reasons.

These benchmarks reflect a significantly more efficient approach compared to traditional drug discovery methods, which often require significantly longer timelines (2.5-4 years) and greater resource expenditure. Insilico's integration of AI and automation streamlines candidate selection and synthesis, demonstrating how innovative approaches can shorten preclinical development timelines.

Benchmarking case studies:

A compelling case study highlighting the benchmarks of Insilico Medicine's developmental candidate package was published earlier in 2024, with a Nature Biotechnology Paper presenting the entire R&D journey from AI algorithms to Phase II clinical trials of ISM001_055, the company's lead drug program with an AI-discovered target and AI-generated design. Following that, Insilico recently announced positive topline results from a Phase IIa trial, where ISM001_055 showed favorable safety and tolerability across all dose levels, as well as dose-dependent response in forced vital capacity (FVC), after only 12 weeks of dosage.

A second case study highlighting the benchmarks set by Insilico Medicine's platform was published in another Nature Biotechnology Paper in December 2024, highlighting the 12 month timeline it took to synthesize and screen approximately 115 molecules, supported by its integrated generative chemistry engine. The paper outlines the early drug discovery and development process and preclinical data of ISM5411. Following that, two separate Phase I studies conducted in Australia and in China indicated that ISM5411 was generally safe and well tolerated in all dose groups, demonstrating a favorable PK profile in validating gut-restrictive properties. 

Insilico Medicine remains steadfast in its commitment to transparency throughout the drug discovery process, recognizing the critical role it plays in advancing global innovation. By openly sharing benchmarks, including developmental candidate timelines and synthesis data, Insilico aims to demonstrate how shedding light on these metrics can drive efficiency across the industry. Transparent reporting not only highlights the capabilities of advanced AI-driven platforms but also underscores the urgency of accelerating the journey from discovery to clinical trials. Ultimately, such efforts enhance global productivity in pharmaceutical development and bring life-saving therapies to patients more rapidly.

New therapeutic areas:

Several recent breakthroughs at Insilico Medicine have led to the development of strong preclinical assets targeting a new focus on pain, obesity and muscle wasting. Multiple preclinical models have shown encouraging results and supported the planning of a next-generation pipeline highlighted by iNAPs - Insilico non-addictive anti-pain therapeutics.

Chronic pain, obesity, and muscle wasting represent significant unmet medical needs, with millions of patients worldwide lacking effective, safe, and accessible treatment options. Current pain management strategies often rely on addictive opioids, contributing to a global epidemic of dependency and overdose. Similarly, therapies for obesity are limited in their efficacy and tolerability, while muscle-wasting conditions often go unaddressed due to a lack of targeted treatments. Insilico Medicine's pivot into these areas signifies a commitment to addressing these critical gaps in care. Leveraging its AI-driven platforms, Insilico aims to rapidly advance the development of novel therapeutics, such as iNAPs, which promise non-addictive, targeted mechanisms of action for pain relief. By integrating cutting-edge computational biology, chemistry, and experimental validation, Insilico seeks to not only accelerate drug discovery timelines but also redefine therapeutic possibilities in these areas, bringing hope to patients with few or no viable options.