Makya

Generative AI for de novo drug design, creating novel, synthetically accessible molecules that rapidly converge toward drug candidates.

Solution by Iktos

Overview

Makya is Iktos's generative AI SaaS platform purpose-built for de novo drug design. Designed by chemists for chemists, it streamlines the drug discovery process by generating diverse, novel, and medchem-like molecules that rapidly converge toward viable drug candidates. Trained on millions of organic reactions, Makya explores a vast chemical space of up to 10²⁷ compounds to deliver optimized molecules faster than traditional approaches.

Makya empowers medicinal chemists and computational chemists alike through an intuitive, easy-to-use interface that facilitates collaboration between medchem and compchem teams. By incorporating Target Product Profile constraints from the outset and leveraging commercial building blocks and organic reactions, the platform places synthetic accessibility at the core of the molecule generation process.

Workflow: How Makya Works

Import your project data and define your Target Product Profile — Makya supports 3D ligand-based and structure-based modeling, SAR modeling, and ADMET property profiling to anchor the generation process in real project requirements.
Explore the infinite possibilities offered by Generative AI — The platform offers a range of generation strategies including growing, linking, cyclization, macrocyclization, and fine tuning, enabling broad and targeted exploration of chemical space.
Select ideal and optimized compounds using all available scores — Researchers can evaluate and prioritize generated molecules based on multi-parametric optimization scores to identify the most promising candidates for synthesis.

Key Capabilities

Faster convergence: Makya accelerates convergence toward a drug candidate by intelligently navigating chemical space with AI-driven optimization.
Multi-parametric optimization: Molecules are generated and ranked based on multiple simultaneous property objectives, ensuring high-value outputs.
Retrosynthesis integration: Synthetic accessibility is built into the generation process, ensuring that designed compounds can realistically be sent to synthesis.
API connectivity: External models and custom scoring functions can be connected via API, providing enhanced flexibility and integration with existing workflows.
New ligand-based design models: The latest Makya release (January 2026) introduces two new generative models that design molecules and their 3D poses directly from a known active ligand — without docking, rigid superposition, or conformer enumeration — keeping the platform extremely fast and efficient even for long, flexible molecules.

Learning and Adoption Resources

Short tutorials, webinars, and publications are available to help users learn Makya at their own pace.
A dedicated demo page provides access to virtual experiment walkthroughs.
Published real-world case studies demonstrate how Makya has delivered faster and more efficient results for drug discovery partners.

Makya is delivered as a SaaS platform, making it accessible without heavy infrastructure requirements. Iktos supports prospective users with demo access, a downloadable product flyer, and a library of scientific publications showcasing the platform's impact across drug discovery projects.