Insilico Medicine

What is Insilico Medicine?

Insilico Medicine is a clinical-stage biotechnology company pioneering the use of generative artificial intelligence for drug discovery and design. Founded in 2014 by Alex Zhavoronkov at Johns Hopkins University, the company develops and commercializes the Pharma.AI suite, an integrated platform covering all stages of pharmacological research: from identifying disease-linked proteins to virtual molecular design and statistical prediction of success in human clinical trials.

How it works

Insilico's Pharma.AI platform is based on deep generative models, such as generative adversarial networks (GANs), reinforcement learning algorithms, and natural language processing. The platform is divided into three functional pillars that operate in an integrated manner:

1. PandaOmics (Biology): Analyzes genomic and proteomic data from millions of patients to identify which specific proteins or genes act as the biological "targets" of a given pathology.
2. Chemistry42 (Chemistry): Creates chemical structures of innovative molecules from scratch that can precisely bind to and inhibit the identified target, while optimizing stability and safety.
3. InClinico (Medicine): Analyzes historical clinical trial data and patient characteristics to predict the probability of approval of new drug candidates in real clinical trials.

Key features

• Omics Target Discovery (PandaOmics): Algorithm that reads academic articles, genetic databases, and patents to map and score ideal therapeutic targets for rare or chronic diseases.
• Generative Chemistry Design (Chemistry42): AI engine that proposes unique chemical formulations ready for physical synthesis in the lab.
• Clinical Trial Prediction (InClinico): Risk assessment that helps investors and pharmaceutical companies understand if a test drug will pass from Phase II to Phase III of regulatory trials.
• Intelligent Robotics Lab: Physical molecular biology laboratory operated entirely by industrial robots and AI to test generated samples autonomously.

Available integrations

As an enterprise scientific solution, the Pharma.AI suite operates in a private cloud integrated with pharmaceutical research supercomputers and global public and private health genomic databases.

Who it is for

• Large pharmaceutical companies seeking to replenish their innovation pipeline with high-impact proprietary molecules.
• Biotechnology institutions that need to reduce the time and cost of early pre-clinical trials.
• Biotech venture capital investors looking to assess the likelihood of success of healthcare treatments in development.

Real-world use cases

1. Development of Fibrosis Treatment: Insilico Medicine designed a new molecular inhibitor to treat Idiopathic Pulmonary Fibrosis. Using AI, the company took only 18 months to advance the compound from initial target discovery to Phase I human clinical trials, a savings of several years and millions of dollars compared to the classic method.
2. Target Identification for ALS: The company used PandaOmics in partnership with academic institutions to map new protein targets linked to Amyotrophic Lateral Sclerosis (ALS), identifying promising genes that had not been previously described.

Pricing

Pros and Cons

Pros:

• End-to-end platform that seamlessly connects omics biology with synthetic molecular chemistry.
• Real track record of AI-designed drugs successfully advancing to human trials.
• The robotic laboratory physically accelerates the validations proposed by artificial intelligence.

Cons:

• Highly complex scientific software, requiring PhDs and molecular biologists for its correct operation.
• Enterprise commercial licensing costs inaccessible to the general public or small clinics.

Alternatives to Insilico Medicine

The main AI alternatives for the pharmacological and biotechnological field are:

• Exscientia: Direct British competitor focusing on molecular design and precision personalized medicine.
• Recursion Pharmaceuticals: Large-scale digitized biological platform focused on cell image mapping assisted by AI.
• Schrödinger: Classic and consolidated software in the global market for physical simulations of molecular interactions.
• BenevolentAI: Focused on crossing biomedical literature data for treatment discovery and drug repositioning.