6.3.1. ilib diverse's filter options
ilib diverse supports various user-adaptable physicochemical property constraints for the generation of compound libraries. Most features are set by defining maximum/minimum values or a standard distribution (determined by mean value and variance).
ilib diverse supports the following constraints for generated molecules:
- Weight: minimum/maximum and standard distribution
- Flexibility: minimum/maximum and standard distribution
- Size: minimum/maximum and standard distribution
- Lipophilicity: minimum/maximum and standard distribution
- Reactivity: on/off
- Chemical features: minimum/maximum
- Atom occurrences: minimum/maximum
The molecular weight setting enables you to control the weight of generated molecules, which in part reflects their size.
Flexibility of molecules represents a major property of drug-like molecules and is characterized by the number of rings and the number of rotatable bonds (i.e. the number of exocyclic single bonds bound to nonterminal heavy atoms; amide C-N bonds are not considered because of their high rotational energy barrier). Rings mostly reduce flexibility, while the number of rotatable bonds raises flexibility crucially resulting in large conformational space. ilib diverse supports all these features for determining the flexibility of drugs and additionally provides the relative number of rotatable bonds feature considering the ligand's size.
Size enables you to control the size of generated molecules by setting the number of aromatic bonds (i.e. the number of bonds participating in aromatic systems), the number of bonds (i.e. the number of covalent bonds) and the number of heavy atoms (i.e. the number all atoms except hydrogen). The size of molecules is determined by adjusting minimum/maximum values or by placing mean and variance for a standard distribution.
The lipophilicity crucially determines i.a. the absorption of drugs. An appropriate setting for lipophilicity is vital for drug action and can be defined by log P (i.e. octanol/water partition coefficient) or the topological surface area (i.e the sum of surface areas of polar atoms like oxygen, nitrogen and attached hydrogen in a molecule). ilib diverse supports both features with an additional relative topological surface area option considering the size of the molecules.
Reactivity is a powerful feature of ilib diverse repelling toxic and reactive groups to be built into compounds. The novel algorithms based on common toxicological knowledge and the analysis of various databases containing drug-like molecules recognize problematic chemical features and reject these compounds.
Chemical features like H-bond acceptors, H-bond donors, positive ionizable and negative ionizable are essential properties determining the ability of ligands to interact with proteins. While H-bond features increase the capability of building up potent interactions with proteins in order to stabilize the complexes, ionic interactions are vital for the approach of drug molecules to the targets.
The atom occurences feature enables you to control the number of nitrogens, oxygens and halogens in generated molecules.