Skip to main content
DS

Denovo Sciences

Denovo Sciences utilizes a reinforcement learning-based platform to design and optimize chemical structures without relying on training datasets, enabling the discovery of novel therapeutics for targets with limited data. This technology allows for the rapid generation of multitarget small molecules, addressing the challenge of complex diseases that require modulation of multiple biological targets simultaneously.

Yerevan, ArmeniaFounded 2019111K+ followers
Updated 20 months ago

Funding

Funding not disclosed

EN
Funding rounds are not available yet.

Founders

Product

Problem

Traditional drug discovery methods often rely on extensive, well-curated training datasets, which are unavailable for many validated molecular targets, limiting the ability to design novel therapeutics. Existing machine learning approaches may also struggle to explore the vast chemical space required to identify novel and synthetically accessible chemical structures, especially for complex diseases requiring multi-target modulation.

Solution

Denovo Sciences offers a reinforcement learning platform that designs and optimizes chemical structures without relying on training datasets, enabling the discovery of novel therapeutics even for targets with limited data. The platform facilitates deep exploration of chemical space, generating thousands of novel and specific chemical structures against the target of interest within hours. It is uniquely designed for the rational design of multi-target chemical structures, accounting for multiple different binding pockets to develop small molecules capable of modulating multiple biological targets simultaneously. This approach addresses the challenges of complex diseases and offers a more predictable pharmacokinetic profile compared to combination therapies.

Target Audience

The primary customers are pharmaceutical companies and research institutions seeking to discover novel therapeutics for targets with limited data or to develop multi-target drugs for complex diseases.

Features

  • Reinforcement learning algorithms for de novo design and optimization of chemical structures independent of training datasets.
  • Rapid generation of thousands of novel, synthetically accessible, and target-specific chemical structures.
  • Target-driven design of chemical structures accounting for multiple binding pockets.
  • Ability to design multi-target small molecules with merged pharmacophores for interaction with multiple proteins.
  • Technology applicable to any type of target, including proteins and nucleic acids.
  • Integration of molecular simulations to address multiple essential pre-clinical conditions.
This profile is AI-generated and may contain inaccuracies.