Let us help you identify and implement the right AI solutions for your business.
The drug discovery process traditionally spans 10-15 years and costs upwards of $2-3 billion per approved drug, with over 90% failure rate in clinical trials due to poor efficacy, toxicity, or ADMET issues.[1] In idiopathic pulmonary fibrosis (IPF), a fatal lung disease with limited treatments like pirfenidone and nintedanib, the need for novel therapies is urgent, but identifying viable targets and designing effective small molecules remains arduous, relying on slow high-throughput screening of existing libraries.[2]
Key challenges include target identification amid vast biological data, de novo molecule generation beyond screened compounds, and predictive modeling of properties to reduce wet-lab failures. Insilico faced skepticism on AI's ability to deliver clinically viable candidates, regulatory hurdles for AI-discovered drugs, and integration of AI with experimental validation.[3]
Insilico deployed its end-to-end Pharma.AI platform, integrating generative AI and deep learning for accelerated discovery. PandaOmics used multimodal deep learning on omics data to nominate novel targets like TNIK kinase for IPF, prioritizing based on disease relevance and druggability.[4]
Chemistry42 employed generative models (GANs, reinforcement learning) to design de novo molecules, generating and optimizing millions of novel structures with desired properties, while InClinico predicted preclinical outcomes. This AI-driven pipeline overcame traditional limitations by virtual screening vast chemical spaces and iterating designs rapidly.[5] Validation through hybrid AI-wet lab approaches ensured robust candidates like ISM001-055 (Rentosertib).[6]
Book a free consultation to explore how AI can solve your specific challenges.
Insilico's implementation began with PandaOmics, a deep learning platform analyzing vast datasets including genomics, transcriptomics, and literature to identify TNIK (TRAF2 and NCK-interacting kinase) as a novel IPF target. Traditional methods miss such targets; AI scored it highly for expression in fibroblasts and druggability, reducing target validation time from years to months.[1][4]
The model integrated transfer learning from protein-protein interactions and pathology data, achieving superior prioritization over rule-based approaches.
Next, Chemistry42 generated de novo small molecules using generative adversarial networks (GANs), variational autoencoders, and reinforcement learning. Starting from TNIK structure, it produced over 400,000 unique scaffolds optimized for potency, selectivity, and ADMET properties like solubility and liver toxicity—properties predicted with deep neural networks trained on millions of compounds.[3][5]
Lead candidate ISM001-055 (Rentosertib) emerged after multi-objective optimization, scoring high in virtual assays simulating fibrosis models. This phase took weeks, versus years for medicinal chemistry iterations.
InClinico employed graph neural networks and transformers to simulate pharmacokinetics, efficacy in humanized models, and safety endpoints. Predictions guided synthesis of top 100 candidates, with 80% matching in vitro results, minimizing lab costs.[2] Wet-lab confirmation showed potent TNIK inhibition (IC50 <1nM), anti-fibrotic effects in IPF organoids, and favorable PK in rodents/non-human primates.
IND filed in November 2021, just 21 months post-project start (Feb 2020). Phase I (2022) confirmed safety in 48 healthy volunteers; Phase II launched July 2023 in China/US for IPF patients, evaluating efficacy via FVC and imaging.[1][6] By 2025, Rentosertib received USAN naming, signaling progress. NVIDIA GPUs powered training, enabling scalability.[5]
Challenges like AI hallucination in molecule feasibility were overcome via hybrid loops: AI proposals synthesized and fed back for retraining. Regulatory buy-in came from transparent data packages and partnerships (e.g., Sanofi). This end-to-end AI pipeline compressed timelines by 80%, positioning Insilico with 30+ programs.[7]
Discover how we can help you implement similar solutions.
Founder & Partner
Reruption GmbH
Falkertstraße 2
70176 Stuttgart
Phone