Chemical Space Navigation

FTrees by Matthias Rarey was among the first tools developed that allowed users to later search for compounds in Chemical Spaces with beyond-billions of molecules. In FTrees, query molecules are translated into fuzzy pharmacophore descriptors. Subsequently those are used to search at unprecended speed for similar molecules in the Chemical Space with a tree alignment approach. Finally, users receive results with numerical descriptors to understand the similarity.
FTrees is part of the Chemical Space navigation platform infiniSee and can be used as standalone commandline tool for programmatic access to the technology. Further information can be found here.

Even a simple task like searching for a specific substructure in Chemical Spaces can become a tedious one due to the sheer size of the space. SpaceMACS is a new method to pull out a defined number of molecules from a space containing a specified substructure — for example the one of a query molecule. The results are the closest chemical analogs allowing fast SAR exploration and compound evolution. In addition, the user receives information on two similarity measures, maximum common substructure-Size (MCS-Size) and MCS-Similarity to allow ranking and comparison of the result molecules.
In summary, SpaceMACS, for the first time, has opened the door to relevant Chemical Space tasks such as identity search, substructure search and substructure-based similarity search. Talk to us for availability.

The next new tool, SpaceLight, utilizes topological fingerprints for Chemical Space navigation to discover close analogs of molecules with high Tanimoto similarity. This has previously not been possible for spaces that exeed the two or three-digit billion size. The new method exploits the combinatorial architecture of fragment-based Chemical Spaces for similarity searches to reliably retrieve similar compound. Although the search can be performed within seconds to minutes on standard hardware, results correlate well with the classical, "non-combinatorial" fingerprint methods, such as ECFP and CSFP.
Contact us if you are interested in testing out SpaceLight for your in-house drug discovery projects.

3D Exploring every corner of a Chemical Space containing billions of enumerated entries necessitates the placement of every ligand into the target with subsequent assessment of the binding quality. Conventional docking would therefore require extensive computational efforts naturally limiting the screening to a few million compounds in most cases. Chemical Space Docking is the workaround solution to access the whole Chemical Space at a much faster rate by initially docking the building blocks of the space with subsequent build-up of the compounds for candidates with a promising binding mode.

Read more about the concept of Chemical Space Docking here.