Which Generator should I use?

Makya provides users with three different generative algorithms that cover various use-cases:

Before starting to generate molecules, please read this section:

Important tips!

• In most cases, a Growing or Linking generator is a good choice for your project:
  • this type of generators can guarantee either novelty or similar molecules, with high quality and always easy synthetic access (as the molecules are built from commercially available building blocks).
• If fragment-based generation does not meet all your criteria (no key fragment to conserve; complex substructure constraints; many goals…), then you can try using the Fine Tuning generator.
  • By playing with the chemical space, you can generate either similar or new molecules. However, generally, the Fine Tuning generator is more conservative than fragment-based generators and less suited to explore a large chemical space.
  • Since the molecules are not built from commercial building blocks, synthetic accessibility is not guaranteed. The retrosynthesis score calculated by Spaya, our retrosynthesis software, can help you filter the generated molecules.

Here are some propositions according to different use cases:

1. Hit discovery from scratch
2. Patent busting
3. Hit expansion, hit-to-lead
4. Lead optimization

Hit discovery from scratch

Useful links:

• Generation of novel ideas when you have no data and no target

Patent busting

Useful links:

• Scaffold Hopping with the Fragment Linking generator

Hit expansion, hit-to-lead

Useful links:

• Growing around a fragment using a 3D reference molecule
• Growing around a fragment using 3D structure-based design
• Hit-To-Lead using Fragment Growing

Fine tuning, lead optimization

Useful links:

• Lead Optimization with the Fine Tuning generator
• Hit-To-Lead using Fragment Growing