SAR and Hit Expansion

SAR study and hit expansion are the next steps of the drug development process after getting the initial hits confirmed. Investigation of related compounds can reveal which parts of the compound are critical (or important) for the activity. Searching for new scaffolds can help identify more active compounds. At Chemspace we offer you different approaches to SAR studies and Hit expansion:

SAR studies

SAR studies schema
  • The hits resulting from Enamine xREAL space are divided into synthons. For each hit, we generate the subset from xREAL by varying only one synthon and fixing the other synthons. This procedure is performed for each synthon in the hit molecule. The generated subsets are docked using ICM-Pro by MolSoft and we select 40-50 compounds for testing.
  • The hits resulting from HTS or DEL campaigns are divided into fragments. For each hit we generate a custom library using the commercially available Building Blocks by varying only one fragment and fixing the other fragments in the hit molecule.
  • Machine learning-driven SAR can be performed for datasets of larger sizes. By building ML models, we can evaluate bit-based (using Morgan fingerprints) or atom-based importance and map them back to the molecules to highlight the substructures that most affect the model’s decision.
  • 2D/3D QSAR permits to uncover correlations between a series of diverse molecular structures and their biological functions for a particular target. To improve classical 2D/3D QSAR modeling we use a combination of 2D/3D QSAR with MolScreen (a hybrid 2D QSAR/fingerprint model kcc(+kca)) from ICM-Pro by Molsoft.
SAR studies schema
Hit Expansion
Similarity Search Schema Hit Expansion studies aim to explore the chemical space around the identified hits and enable scaffold core-hopping techniques which can be used to generate new hit series with improved properties.
  • 2D Similarity (ECFP4 fingerprints) in Chemspace in-stock collection and Enamine REAL Database
  • SpaceLight similarity in Enamine xREAL Space*. SpaceLight utilizes topological fingerprints for xREAL Space navigation to discover close analogs of molecules with high Tanimoto similarity.
  • Fuzzy similarity (FTrees) in Enamine xREAL Space*. The query molecules are translated into fuzzy pharmacophore descriptors which are used to search for similar molecules in the xREAL Space with a tree alignment approach. This method can be used for scaffold hopping.
  • SpaceMACS Search in Enamine xREAL Space*. SpaceMACS is a method to pull out a defined number of molecules from xREAL Space containing a specified substructure. Ranking and comparison of the resulting molecules are done with maximum common substructure-Size (MCS-Size) and MCS-Similarity metrics.
  • Custom Space generation based on similar synthons from Enamine xREAL Space. For each synthon/fragment in the hit molecule we run similarity search to find closest synthons from MEL (Minimal Enumeration Library, 6M compounds). The high-scoring synthons undergo enumeration and docking.
  • 3D Pharmacophore Similarity using RIDE by MolSoft. Performed using Atomic Property Fields (APF) - a grid 3D pharmacophore potential that is generated from one or more high-affinity scaffolds with properties assigned from the empiric physico-chemical components.
* These types of searches are provided by BioSolveIT.
Similarity Search Schema
For prioritization of the compounds for testing we are using the following approaches:
  • Virtual screening approaches
  • Steered MD
  • 2D/3Q QSAR models
  • ML-assisted models trained on HTS or DEL data
  • ML-assisted models for ADME properties prediction.
Set of compounds selected for synthesis Set of compounds selected for synthesis

Would you like to get more details?

Fill out the form to contact our team
Please fill out this field
Please enter a valid email address
Please fill out this field
Please fill out this field
Please fill out this field
Thank you, we will contact you shortly
An error occured