LabGenius
Accelerating protein-based drug discovery through an autonomous, closed-loop robotic platform.

AI-driven de novo drug design and automated retrosynthesis for accelerated medicinal chemistry.
Iktos is a leading French AI company specializing in deep learning applications for medicinal chemistry and drug design. As of 2026, Iktos has solidified its position as a top-tier provider of generative modeling solutions for de novo design (Makya) and AI-powered retrosynthesis (Spaya). Their technical architecture leverages advanced Deep Generative Models, including Reinforcement Learning and Variational Autoencoders, to navigate vast chemical spaces while optimizing for multiple parameters simultaneously (MPO). Unlike traditional docking-based virtual screening, Iktos focuses on generating synthetically accessible molecules that meet specific ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) profiles. The 2026 market position of Iktos is defined by its 'Iktos Robotics' integration—a closed-loop platform that combines generative software with automated chemical synthesis hardware. This synergy addresses the primary bottleneck in drug discovery: the cycle time between molecular design and physical testing. Their platform is utilized by global pharmaceutical giants and mid-sized biotech firms to reduce the 'Design-Make-Test-Analyze' cycle from months to weeks, utilizing proprietary algorithms that ensure high synthetic feasibility and structural diversity.
Uses Reinforcement Learning to optimize molecular structures against custom scoring functions.
Accelerating protein-based drug discovery through an autonomous, closed-loop robotic platform.
Deciphering the gut-brain axis through AI-driven drug discovery for transformative therapeutics.
Accelerating drug discovery through an end-to-end generative AI pipeline for target identification, molecular design, and clinical trial prediction.
Engineering biology at scale to discover and develop next-generation therapeutics.
Verified feedback from the global deployment network.
Post queries, share implementation strategies, and help other users.
A template-based and template-free hybrid AI for predicting chemical reaction pathways.
Weighted scoring system for balancing conflicting properties like solubility vs. potency.
Incorporation of 3D protein structure information into the generative reward function.
Algorithms designed to replace a core molecular scaffold while maintaining peripheral interactions.
Models that improve iteratively as new experimental data is uploaded into the system.
Direct communication between generative software and liquid-handling robots.
Identifying novel chemical matter when traditional screening libraries fail.
Registry Updated:2/7/2026
Purchase building blocks for top candidates
Improving the potency and ADMET profile of a promising chemical lead.
Finding novel chemical structures that bypass existing patents.