Makya

Overview

Makya is a generative AI SaaS platform developed by Iktos for medicinal and computational chemistry teams working on de novo drug design. Its role is to propose molecules against a project’s Target Product Profile and help teams explore, compare, and prioritize candidates. The supplied product page describes ligand- and structure-based design, SAR modeling, and ADMET properties as elements of the project workflow, with multiple objectives considered during molecular generation.

The described workflow starts with importing project data and defining design objectives. Molecular design operations include growing, linking, cyclization, macrocyclization, and fine tuning. Synthesis constraints are incorporated through commercial building blocks and organic reactions, while retrosynthesis is presented as part of the platform. Outputs include generated candidate molecules and associated scores or optimized attributes; the page also describes ligand-based models that generate molecules and their 3D poses from a known active ligand.

For candidate assessment, Makya provides chemical-space exploration, compound insights relative to a reference ligand, and filters using custom scores and 3D interactions. An API enables connections to external models and scoring functions. These are vendor-described capabilities, not independently established performance results. The supplied material does not specify accepted file formats, export schemas, API authentication, or access terms. Its claims about novelty, convergence, and synthetic feasibility should therefore be treated as evaluation questions rather than guarantees for a particular discovery project.

Key Features

• Molecular generation guided by Target Product Profile constraints and multi-parametric optimization.
• Synthesis-aware design using commercial building blocks and organic reactions, with retrosynthesis incorporated into the platform.
• Growing, linking, cyclization, macrocyclization, and fine-tuning operations for molecular design.
• Ligand- and structure-based modeling, including described ligand-based models that generate molecules and 3D poses from a known active ligand.
• Chemical-space exploration, reference-ligand comparisons, and candidate filtering by custom scores and 3D interactions.
• API connections for external models and scoring functions.

Use Cases

• Intended evaluation: generate candidate series against a project’s Target Product Profile and assess how well the proposals balance competing molecular objectives.
• Intended evaluation: explore growing, linking, or macrocyclization while checking whether the resulting candidates meet project-specific synthesis constraints.
• Intended evaluation: use a known active ligand to explore molecular designs and proposed 3D poses, then assess their relevance to the project.
• Intended evaluation: connect an external scoring model through the API and examine how its scores support candidate filtering and prioritization.

How to Use

1. Read the official Makya product page to identify the design modes relevant to your project. Use the supplied description as a capability outline, not an input-format specification.
2. Consult the linked Makya flyer and release factsheet for follow-up product information; their contents were not supplied here.
3. Request a demonstration through Iktos contact. Ask about access terms, supported imports and exports, data handling, and API requirements before supplying project data.
4. For an intended evaluation, define your Target Product Profile, reference ligand where relevant, and synthesis constraints. The documented workflow begins with importing project data and setting these objectives.
5. During the evaluation, explore candidates using the relevant design operations, scores, and 3D-interaction filters. Assess proposed compounds against your own modeling and synthesis criteria rather than assuming vendor claims establish project-specific suitability.